PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise

Advanced search

Message boards : Generalized Fermat Prime Search : Genefer OCL 4

Author Message
Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90626 - Posted: 21 Dec 2015 | 15:09:13 UTC

A new Genefer application "geneferocl4" is available (32-bit binary for windows) from repository https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/windows.

geneferocl4 transform is a NTT (like ocl3).
OCL3 residue numeral system is a single 64-bit prime (2^64 - 2^32 + 1).
OCL4 RNS is a set of two 32-bit primes (3*2^30 + 1 and 13*2^28 + 1).

bMax is 28% smaller than bMax of ocl3.

OCL4 runs faster than OCL3 on a GTX 680: Genefer Mark = 57 / 47 but is slower on a GTX 780 : 61 / 73.

I'm interested in benchmarks. I expect that it will run faster on ATI...?

Thanks, Yves

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90628 - Posted: 21 Dec 2015 | 15:34:09 UTC
Last modified: 21 Dec 2015 | 15:54:59 UTC

Benchmarks for a Gtx970, OCed to 1490mhz. Apparently, OCL 3 is still king. OCL 4 seems to be about as good as the original OCL.

OCL: 78
OCL2: 30
OCL3: 91
OCL4: 80

Command line: geneferocl_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL" Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '359.06'. 2199064^8192+1 Time: 43.6 us/mul. Err: 0.1875 51956 digits 1798620^16384+1 Time: 46.9 us/mul. Err: 0.1875 102481 digits 1471094^32768+1 Time: 65.4 us/mul. Err: 0.1875 202102 digits 1203210^65536+1 Time: 91.9 us/mul. Err: 0.1875 398482 digits 984108^131072+1 Time: 171 us/mul. Err: 0.1797 785521 digits 804904^262144+1 Time: 336 us/mul. Err: 0.1719 1548156 digits 658332^524288+1 Time: 655 us/mul. Err: 0.1797 3050541 digits 538452^1048576+1 Time: 1.31 ms/mul. Err: 0.1738 6009544 digits 440400^2097152+1 Time: 2.57 ms/mul. Err: 0.1777 11836006 digits 360204^4194304+1 Time: 5.35 ms/mul. Err: 0.1592 23305854 digits 294612^8388608+1 Time: 10.6 ms/mul. Err: 0.1797 45879398 digits Genefer Mark = 78. Priority change succeeded. ... Command line: geneferocl2_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL2" Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '359.06'. 2199064^8192+1 Time: 112 us/mul. Err: 0.0001 51956 digits 1798620^16384+1 Time: 119 us/mul. Err: 0.0001 102481 digits 1471094^32768+1 Time: 130 us/mul. Err: 0.0001 202102 digits 1203210^65536+1 Time: 198 us/mul. Err: 0.0001 398482 digits 984108^131072+1 Time: 483 us/mul. Err: 0.0001 785521 digits 804904^262144+1 Time: 859 us/mul. Err: 0.0001 1548156 digits 658332^524288+1 Time: 1.84 ms/mul. Err: 0.0001 3050541 digits 538452^1048576+1 Time: 3.69 ms/mul. Err: 0.0001 6009544 digits 440400^2097152+1 Time: 6.91 ms/mul. Err: 0.0001 11836006 digits 360204^4194304+1 Time: 12.7 ms/mul. Err: 0.0001 23305854 digits 294612^8388608+1 Time: 28 ms/mul. Err: 0.0001 45879398 digits Genefer Mark = 30. Priority change succeeded. ... Command line: geneferocl3_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL3" Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '359.06'. 2199064^8192+1 Time: 64.4 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 57.3 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 64.7 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 85.3 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 162 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 303 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 585 us/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 1.14 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 2.19 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 4.48 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 9.15 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 91. Priority change succeeded. ... Command line: geneferocl4_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '359.06'. 2199064^8192+1 Time: 60 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 64.5 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 69.2 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 95 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 186 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 346 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 647 us/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 1.25 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 2.56 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 5.16 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 10.5 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 80. Priority change succeeded.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90631 - Posted: 21 Dec 2015 | 16:03:35 UTC
Last modified: 21 Dec 2015 | 16:04:23 UTC

And benchmarks for the AMD A6-3500 iGPU. OCL1 straight up doesn't work, and I bet OCL2's score is even lower than 1. Regardless, OCL4 seems like a small speedup from OCL3, so that's good..

OCL: -
OCL2: 1
OCL3: 4
OCL4: 5

Command line: geneferocl2_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL2" Running on platform 'AMD Accelerated Parallel Processing', device 'BeaverCreek', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 908 us/mul. Err: 0.0001 51956 digits 1798620^16384+1 Time: 1.89 ms/mul. Err: 0.0001 102481 digits 1471094^32768+1 Time: 3.96 ms/mul. Err: 0.0001 202102 digits 1203210^65536+1 Time: 2.84 ms/mul. Err: 0.0001 398482 digits 984108^131072+1 Time: 6.41 ms/mul. Err: 0.0001 785521 digits 804904^262144+1 Time: 15.6 ms/mul. Err: 0.0001 1548156 digits 658332^524288+1 Time: 56.9 ms/mul. Err: 0.0001 3050541 digits 538452^1048576+1 Time: 142 ms/mul. Err: 0.0001 6009544 digits 440400^2097152+1 Time: 288 ms/mul. Err: 0.0001 11836006 digits 360204^4194304+1 Time: 224 ms/mul. Err: 0.0001 23305854 digits 294612^8388608+1 Time: 486 ms/mul. Err: 0.0001 45879398 digits Genefer Mark = 1. Priority change succeeded. ... Command line: geneferocl3_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL3" Running on platform 'AMD Accelerated Parallel Processing', device 'BeaverCreek', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 219 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 388 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 742 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 1.5 ms/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 3.06 ms/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 6.3 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 12.8 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 26.7 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 54.6 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 115 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 232 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 4. Priority change succeeded. ... Command line: geneferocl4_windows -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'AMD Accelerated Parallel Processing', device 'BeaverCreek', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 209 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 335 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 631 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 1.26 ms/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 2.55 ms/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 5.1 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 10.5 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 21.7 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 44.3 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 91.2 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 183 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 5. Priority change succeeded.

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90632 - Posted: 21 Dec 2015 | 16:15:44 UTC - in response to Message 90626.

R9 280x
VNC won't let me cut and paste (when I get on the console I can grab them), so in sum:

OCL: 132
OCL2: 13
OCL3: 54
OCL4: 61

Profile maarten1001
Avatar
Send message
Joined: 6 May 07
Posts: 8
ID: 7935
Credit: 25,300,895
RAC: 0
321 LLR Silver: Earned 100,000 credits (119,822)Cullen LLR Silver: Earned 100,000 credits (146,606)ESP LLR Silver: Earned 100,000 credits (105,813)PPS LLR Silver: Earned 100,000 credits (242,825)PSP LLR Silver: Earned 100,000 credits (312,489)SoB LLR Silver: Earned 100,000 credits (112,836)SR5 LLR Silver: Earned 100,000 credits (244,319)SGS LLR Silver: Earned 100,000 credits (329,222)TRP LLR Silver: Earned 100,000 credits (398,190)Woodall LLR Silver: Earned 100,000 credits (162,514)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (276,616)PPS Sieve Jade: Earned 10,000,000 credits (11,694,533)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (103,064)TRP Sieve (suspended) Gold: Earned 500,000 credits (506,564)GFN Turquoise: Earned 5,000,000 credits (5,912,194)PSA Ruby: Earned 2,000,000 credits (4,631,751)
Message 90633 - Posted: 21 Dec 2015 | 16:59:25 UTC

GTX580

OCL: 77
OCL2: 20
OCL3: 40
OCL4: 47

GTX460

OCL: 31
OCL2: 8
OCL3: 19
OCL4: 26

Profile TylerProject donor
Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,280,170,555
RAC: 7,054
Discovered 1 mega primeFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (5,238,816)Cullen LLR Turquoise: Earned 5,000,000 credits (5,124,261)ESP LLR Turquoise: Earned 5,000,000 credits (8,479,525)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,656,837)PPS LLR Sapphire: Earned 20,000,000 credits (23,667,318)PSP LLR Sapphire: Earned 20,000,000 credits (37,386,303)SoB LLR Sapphire: Earned 20,000,000 credits (21,004,651)SR5 LLR Sapphire: Earned 20,000,000 credits (33,215,655)SGS LLR Turquoise: Earned 5,000,000 credits (5,000,408)TRP LLR Jade: Earned 10,000,000 credits (12,199,250)Woodall LLR Turquoise: Earned 5,000,000 credits (8,449,472)321 Sieve Turquoise: Earned 5,000,000 credits (8,012,059)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (27,921,016)PPS Sieve Double Gold: Earned 500,000,000 credits (542,336,593)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (7,990,608)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,217,057)AP 26/27 Turquoise: Earned 5,000,000 credits (7,111,637)GFN Sapphire: Earned 20,000,000 credits (31,676,706)PSA Double Silver: Earned 200,000,000 credits (420,586,383)
Message 90637 - Posted: 21 Dec 2015 | 17:56:55 UTC

GTX 760 Stock:
OCL: 55
OCL2: 12
OCL3: 38
OCL4: 48
____________

275*2^3585539+1 is prime!!! (1079358 digits)

Proud member of Aggie the Pew

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2460
ID: 29980
Credit: 442,802,854
RAC: 10,291
Discovered 4 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (8,774,878)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (6,454,573)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,122,074)PPS LLR Emerald: Earned 50,000,000 credits (83,377,417)PSP LLR Jade: Earned 10,000,000 credits (15,223,714)SoB LLR Jade: Earned 10,000,000 credits (17,319,914)SR5 LLR Sapphire: Earned 20,000,000 credits (23,996,561)SGS LLR Turquoise: Earned 5,000,000 credits (7,342,780)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Jade: Earned 10,000,000 credits (19,866,589)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (21,918,894)GFN Emerald: Earned 50,000,000 credits (76,466,089)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 90638 - Posted: 21 Dec 2015 | 18:19:54 UTC

980Ti results:

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 980 Ti', version 'OpenCL 1.2 CUDA' and driver '359.06'. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" 2199064^8192+1 Time: 79.1 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 80.4 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 85.2 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 94 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 164 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 278 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 518 us/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 996 us/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 1.99 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 3.81 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 8.11 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 104. Priority change succeeded.


For OCL3 I get 119 marks, so OCL4 is slower in this case.

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2460
ID: 29980
Credit: 442,802,854
RAC: 10,291
Discovered 4 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (8,774,878)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (6,454,573)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,122,074)PPS LLR Emerald: Earned 50,000,000 credits (83,377,417)PSP LLR Jade: Earned 10,000,000 credits (15,223,714)SoB LLR Jade: Earned 10,000,000 credits (17,319,914)SR5 LLR Sapphire: Earned 20,000,000 credits (23,996,561)SGS LLR Turquoise: Earned 5,000,000 credits (7,342,780)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Jade: Earned 10,000,000 credits (19,866,589)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (21,918,894)GFN Emerald: Earned 50,000,000 credits (76,466,089)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 90639 - Posted: 21 Dec 2015 | 18:32:39 UTC - in response to Message 90632.

R9 280x
VNC won't let me cut and paste (when I get on the console I can grab them), so in sum:

OCL: 132
OCL2: 13
OCL3: 54
OCL4: 61


Same problem here, with similar scores on my R9 280X.

OCL3: 55
OCL4: 59

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90640 - Posted: 21 Dec 2015 | 19:11:49 UTC

HD7950 w/Boost
OCL: 109
OCL2: 11
OCL3: 44
OCL4: 50

Command line: geneferocl_windows.exe -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL" Running on platform 'AMD Accelerated Parallel Processing', device 'Tahiti', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 120 us/mul. Err: 0.1875 51956 digits 1798620^16384+1 Time: 126 us/mul. Err: 0.1875 102481 digits 1471094^32768+1 Time: 108 us/mul. Err: 0.1875 202102 digits 1203210^65536+1 Time: 94.8 us/mul. Err: 0.1875 398482 digits 984108^131072+1 Time: 140 us/mul. Err: 0.1797 785521 digits 804904^262144+1 Time: 412 us/mul. Err: 0.1719 1548156 digits 658332^524288+1 Time: 548 us/mul. Err: 0.1797 3050541 digits 538452^1048576+1 Time: 1.02 ms/mul. Err: 0.1758 6009544 digits 440400^2097152+1 Time: 1.85 ms/mul. Err: 0.1680 11836006 digits 360204^4194304+1 Time: 3.62 ms/mul. Err: 0.1563 23305854 digits 294612^8388608+1 Time: 7.55 ms/mul. Err: 0.1719 45879398 digits Genefer Mark = 109. ------------------------------- Command line: geneferocl2_windows.exe -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL2" Running on platform 'AMD Accelerated Parallel Processing', device 'Tahiti', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 389 us/mul. Err: 0.0001 51956 digits 1798620^16384+1 Time: 388 us/mul. Err: 0.0001 102481 digits 1471094^32768+1 Time: 698 us/mul. Err: 0.0001 202102 digits 1203210^65536+1 Time: 874 us/mul. Err: 0.0001 398482 digits 984108^131072+1 Time: 1.91 ms/mul. Err: 0.0001 785521 digits 804904^262144+1 Time: 3.73 ms/mul. Err: 0.0001 1548156 digits 658332^524288+1 Time: 7.9 ms/mul. Err: 0.0001 3050541 digits 538452^1048576+1 Time: 16.3 ms/mul. Err: 0.0001 6009544 digits 440400^2097152+1 Time: 22.1 ms/mul. Err: 0.0001 11836006 digits 360204^4194304+1 Time: 24.4 ms/mul. Err: 0.0001 23305854 digits 294612^8388608+1 Time: 87.1 ms/mul. Err: 0.0001 45879398 digits Genefer Mark = 11. ----------------------------- Command line: geneferocl3_windows.exe -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL3" Running on platform 'AMD Accelerated Parallel Processing', device 'Tahiti', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 124 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 183 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 250 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 309 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 362 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 676 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 1.23 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 2.34 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 4.67 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 9.37 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 18.8 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 44. ----------------- Command line: geneferocl4_windows.exe -b Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'AMD Accelerated Parallel Processing', device 'Tahiti', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. 2199064^8192+1 Time: 170 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 122 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 202 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 242 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 342 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 656 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 1.14 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 2.08 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 4.07 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 8.18 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 16.8 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 50.


It's a slight improvement over OCL3, which is nice. I don't know the timeline, but with a possible 4 OCL types out now, what's the progress on a single "choose the fastest" program?
____________
Eating more cheese on Thursdays.

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 90641 - Posted: 21 Dec 2015 | 19:14:14 UTC

Fury Nano

OCL3: 70
OCL4: 86


Running benchmarks for transform implementation "OCL3" Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'. 2199064^8192+1 Time: 100 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 109 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 114 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 145 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 225 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 350 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 794 us/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 1.5 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 2.91 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 5.89 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 11.8 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 70. Priority change succeeded. Running benchmarks for transform implementation "OCL4" Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'. 2199064^8192+1 Time: 108 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 88.6 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 98.1 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 121 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 216 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 311 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 653 us/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 1.25 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 2.39 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 4.71 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 9.21 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 86. Priority change succeeded.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90642 - Posted: 21 Dec 2015 | 19:28:13 UTC

GT430 w/DDR3.

OCL: 8
OCL2: 2
OCL3: 5
OCL4: 7

Not a bad improvement, but I wouldn't run GFN on this GPU at those speeds.

Command line: geneferocl4_windows.exe -b -d 1 Priority change succeeded. Priority change succeeded. Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'NVIDIA CUDA', device 'GeForce GT 430', version 'OpenCL 1.1 CUDA' and driver '361.43'. 2199064^8192+1 Time: 116 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 220 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 396 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 798 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 1.7 ms/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 3.46 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 7.39 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 15 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 29.7 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 62.1 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 137 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 7.

____________
Eating more cheese on Thursdays.

ich_eben*Project donor
Volunteer tester
Avatar
Send message
Joined: 14 Oct 10
Posts: 101
ID: 70063
Credit: 306,880,766
RAC: 1,581
321 LLR Sapphire: Earned 20,000,000 credits (20,143,284)Cullen LLR Jade: Earned 10,000,000 credits (11,151,369)ESP LLR Jade: Earned 10,000,000 credits (10,492,204)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (10,506,320)PPS LLR Jade: Earned 10,000,000 credits (15,287,978)PSP LLR Jade: Earned 10,000,000 credits (11,779,216)SoB LLR Jade: Earned 10,000,000 credits (11,538,563)SR5 LLR Sapphire: Earned 20,000,000 credits (21,832,970)SGS LLR Jade: Earned 10,000,000 credits (10,182,339)TRP LLR Jade: Earned 10,000,000 credits (11,478,625)Woodall LLR Jade: Earned 10,000,000 credits (10,382,751)321 Sieve Sapphire: Earned 20,000,000 credits (30,072,122)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,023,474)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (51,836,409)PPS Sieve Sapphire: Earned 20,000,000 credits (20,426,047)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,033,433)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,023,749)AP 26/27 Ruby: Earned 2,000,000 credits (3,129,282)GFN Jade: Earned 10,000,000 credits (18,187,599)PSA Jade: Earned 10,000,000 credits (14,365,706)
Message 90647 - Posted: 21 Dec 2015 | 21:48:09 UTC

GTX 670

OCL 62
OCL2 14
OCL3 42
OCL4 48


Running benchmarks for transform implementation "OCL"

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 670', version 'OpenCL 1.1
CUDA' and driver '344.11'.

2199064^8192+1 Time: 45.6 us/mul. Err: 0.1875 51956 digits
1798620^16384+1 Time: 46.4 us/mul. Err: 0.1875 102481 digits
1471094^32768+1 Time: 63.7 us/mul. Err: 0.1875 202102 digits
1203210^65536+1 Time: 112 us/mul. Err: 0.1875 398482 digits
984108^131072+1 Time: 217 us/mul. Err: 0.1797 785521 digits
804904^262144+1 Time: 394 us/mul. Err: 0.1719 1548156 digits
658332^524288+1 Time: 768 us/mul. Err: 0.1797 3050541 digits
538452^1048576+1 Time: 1.57 ms/mul. Err: 0.1738 6009544 digits
440400^2097152+1 Time: 3.24 ms/mul. Err: 0.1777 11836006 digits
360204^4194304+1 Time: 6.75 ms/mul. Err: 0.1592 23305854 digits
294612^8388608+1 Time: 14.5 ms/mul. Err: 0.1797 45879398 digits
Genefer Mark = 62.
Priority change succeeded.

Running benchmarks for transform implementation "OCL2"

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 670', version 'OpenCL 1.1
CUDA' and driver '344.11'.

2199064^8192+1 Time: 95.7 us/mul. Err: 0.0001 51956 digits
1798620^16384+1 Time: 109 us/mul. Err: 0.0001 102481 digits
1471094^32768+1 Time: 210 us/mul. Err: 0.0001 202102 digits
1203210^65536+1 Time: 444 us/mul. Err: 0.0001 398482 digits
984108^131072+1 Time: 842 us/mul. Err: 0.0001 785521 digits
804904^262144+1 Time: 1.6 ms/mul. Err: 0.0001 1548156 digits
658332^524288+1 Time: 3.32 ms/mul. Err: 0.0001 3050541 digits
538452^1048576+1 Time: 6.91 ms/mul. Err: 0.0001 6009544 digits
440400^2097152+1 Time: 15.1 ms/mul. Err: 0.0001 11836006 digits
360204^4194304+1 Time: 29.7 ms/mul. Err: 0.0001 23305854 digits
294612^8388608+1 Time: 59.7 ms/mul. Err: 0.0001 45879398 digits
Genefer Mark = 14.
Priority change succeeded.

Running benchmarks for transform implementation "OCL3"

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 670', version 'OpenCL 1.1
CUDA' and driver '344.11'.

2199064^8192+1 Time: 50.2 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 58.4 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 85.6 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 152 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 303 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 588 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 1.17 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 2.41 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 4.69 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 9.87 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 20.9 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 42.
Priority change succeeded.

Running benchmarks for transform implementation "OCL4"

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 670', version 'OpenCL 1.1
CUDA' and driver '344.11'.

2199064^8192+1 Time: 44.4 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 49.3 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 79 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 134 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 267 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 502 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 1.03 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 2.01 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 4.15 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 8.68 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 18.1 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 48.
Priority change succeeded.

____________

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90649 - Posted: 22 Dec 2015 | 0:39:58 UTC - in response to Message 90626.

Titan Black
OCL 130
OCl2 32
OCL3 84
OCL4 75


Command line: geneferocl_windows.exe -b

Priority change succeeded.
Priority change succeeded.
Generalized Fermat Number Bench
Running benchmarks for transform implementation "OCL"

Running on platform 'NVIDIA CUDA', device 'GeForce GTX TITAN Black', version 'OpenCL 1.2 CUDA' and driver '359.00'.

2199064^8192+1 Time: 59.4 us/mul. Err: 0.1875 51956 digits
1798620^16384+1 Time: 59.6 us/mul. Err: 0.1875 102481 digits
1471094^32768+1 Time: 71.2 us/mul. Err: 0.1875 202102 digits
1203210^65536+1 Time: 81.7 us/mul. Err: 0.1875 398482 digits
984108^131072+1 Time: 133 us/mul. Err: 0.1797 785521 digits
804904^262144+1 Time: 239 us/mul. Err: 0.1719 1548156 digits
658332^524288+1 Time: 421 us/mul. Err: 0.1797 3050541 digits
538452^1048576+1 Time: 790 us/mul. Err: 0.1738 6009544 digits
440400^2097152+1 Time: 1.56 ms/mul. Err: 0.1777 11836006 digits
360204^4194304+1 Time: 3.11 ms/mul. Err: 0.1592 23305854 digits
294612^8388608+1 Time: 6.49 ms/mul. Err: 0.1797 45879398 digits
Genefer Mark = 130.
Priority change succeeded.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90653 - Posted: 22 Dec 2015 | 9:55:29 UTC

Thanks.

Then OCL4 is faster than OCL3 except for GK110 and Maxwell (CC 3.5 and 5.0).

The "funnel shift" instruction is the most likely reason: i = sqrt(-1) = 2^48 (mod 2^64 - 2^32 + 1).
x * i = x << 48 (mod 2^64 - 2^32 + 1).
This trick doesn't work with 3*2^30 + 1 or 13*2^28 + 1.

Profile DeleteNullProject donor
Volunteer tester
Avatar
Send message
Joined: 6 Apr 06
Posts: 242
ID: 2663
Credit: 7,255,283,251
RAC: 126,806
Discovered 26 mega primesEliminated 3 conjecture "k"sFound 4 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 2 mega primes in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 2 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Emerald: Earned 50,000,000 credits (60,657,554)Cullen LLR Emerald: Earned 50,000,000 credits (67,814,375)ESP LLR Emerald: Earned 50,000,000 credits (61,429,176)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (59,199,415)PPS LLR Double Silver: Earned 200,000,000 credits (246,743,264)PSP LLR Emerald: Earned 50,000,000 credits (76,449,391)SoB LLR Emerald: Earned 50,000,000 credits (71,933,705)SR5 LLR Double Bronze: Earned 100,000,000 credits (102,395,858)SGS LLR Emerald: Earned 50,000,000 credits (54,295,707)TPS LLR (retired) Bronze: Earned 10,000 credits (61,785)TRP LLR Double Bronze: Earned 100,000,000 credits (102,088,428)Woodall LLR Emerald: Earned 50,000,000 credits (50,544,668)321 Sieve Emerald: Earned 50,000,000 credits (63,039,849)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,685,309)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (53,324,439)PPS Sieve Double Ruby: Earned 2,000,000,000 credits (3,787,521,227)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,057,785)TRP Sieve (suspended) Emerald: Earned 50,000,000 credits (50,146,560)AP 26/27 Double Silver: Earned 200,000,000 credits (254,699,080)GFN Double Amethyst: Earned 1,000,000,000 credits (1,799,551,280)PSA Double Silver: Earned 200,000,000 credits (204,786,115)
Message 90657 - Posted: 22 Dec 2015 | 10:27:31 UTC - in response to Message 90653.

AMD 390X:
OCL 139
OCL2 20
OCL3 82
OCL4 90

Titan Black:
OCL 128
OCL2 32
OCL3 83
OCL4 74

____________
DeleteNull

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 90693 - Posted: 23 Dec 2015 | 17:10:44 UTC

Tested on my GTX 580:

OCL3: 42
OCL4: 48

Looking at the speeds on each n is sort of interesting:

n gain 15 5.28% 16 22.01% *** 17 8.36% 18 8.17% 19 9.57% 20 12.07% *** 21 11.13% 22 13.30%


N=16 (65536) and n=20 (1048576) look like the best places to use OCL4 right now, at least for me. N=15 not only show a smaller speed increase, but it's beyond OCL4's (and OCL3's) b limit. N=21 and 22 are currently using OCL, which on my GPU is substantially faster.

Those numbers are repeatable. N=16 seems out of place, but that's what the benchmarks are indicating. It's not just the benchmarks, either. Actual computations show the same speed increase. OCL3 runs the n=16 tests in 4 minutes while OCL4 runs them in 3 minutes.

____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90695 - Posted: 23 Dec 2015 | 17:41:01 UTC - in response to Message 90693.

Tested on my GTX 580:

OCL3: 42
OCL4: 48

Looking at the speeds on each n is sort of interesting:

n gain 15 5.28% 16 22.01% *** 17 8.36% 18 8.17% 19 9.57% 20 12.07% *** 21 11.13% 22 13.30%


N=16 (65536) and n=20 (1048576) look like the best places to use OCL4 right now, at least for me. N=15 not only show a smaller speed increase, but it's beyond OCL4's (and OCL3's) b limit. N=21 and 22 are currently using OCL, which on my GPU is substantially faster.

Those numbers are repeatable. N=16 seems out of place, but that's what the benchmarks are indicating. It's not just the benchmarks, either. Actual computations show the same speed increase. OCL3 runs the n=16 tests in 4 minutes while OCL4 runs them in 3 minutes.

Maybe it's because OCL4 is "a little" more efficient at higher N, but "A LOT" more efficient at higher B. n=16 is almost at 5M, while n=17~22 are around 2M and below. Maybe. Perhaps if you try 5M on n=17 with with both OCL3 and 4 (assuming 4 can go that high), you'll see a similar % boost.

Although my 970 is actually slower with the new app, it seems that for B=2M OCL4 is 10% slower, but for B=5M, it's only 9.3% slower. With that said, I wonder how non-Maxwell cards perform....

Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 90735 - Posted: 28 Dec 2015 | 18:03:33 UTC - in response to Message 90626.
Last modified: 28 Dec 2015 | 18:03:51 UTC

I uploaded "geneferocl4" for Linux 32/64. Both apps can be downloaded from https://www.assembla.com/spaces/genefer/subversion/commits/964
____________
Best wishes. Knowledge is power. by jjwhalen

Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 90736 - Posted: 28 Dec 2015 | 18:22:22 UTC - in response to Message 90735.

geneferocl4_linux32

boinc@vmware2k-3:~/$ ./geneferocl4_linux32 -b geneferocl4 3.2.10-dev (Linux/OpenCL-NTT-32-2/32-bit) Copyright 2001-2015, Yves Gallot Copyright 2009, Mark Rodenkirch, David Underbakke Copyright 2010-2012, Shoichiro Yamada, Ken Brazier Copyright 2011-2015, Iain Bethune, Michael Goetz, Ronald Schneider Genefer is free source code, under the MIT license. Command line: ./geneferocl4_linux32 -b Priority change succeeded. Priority change failed (needs superuser privileges). Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'NVIDIA CUDA', device 'GeForce GTS 450', version 'OpenCL 1.1 CUDA' and driver '352.63'. 2199064^8192+1 Time: 87.5 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 114 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 212 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 448 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 928 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.83 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 3.81 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 7.6 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 15.5 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 32.5 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 71.4 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 13. Priority change succeeded.




geneferocl4_linux64

boinc@vmware2k-3:~/$ ./geneferocl4_linux64 -b geneferocl4 3.2.10-dev (Linux/OpenCL-NTT-32-2/64-bit) Copyright 2001-2015, Yves Gallot Copyright 2009, Mark Rodenkirch, David Underbakke Copyright 2010-2012, Shoichiro Yamada, Ken Brazier Copyright 2011-2015, Iain Bethune, Michael Goetz, Ronald Schneider Genefer is free source code, under the MIT license. Command line: ./geneferocl4_linux64 -b Priority change succeeded. Priority change failed (needs superuser privileges). Generalized Fermat Number Bench Running benchmarks for transform implementation "OCL4" Running on platform 'NVIDIA CUDA', device 'GeForce GTS 450', version 'OpenCL 1.1 CUDA' and driver '352.63'. 2199064^8192+1 Time: 67.3 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 121 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 239 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 493 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 990 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.92 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 3.95 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 8.02 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 17 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 35.4 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 76.1 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 12. Priority change succeeded.

____________
Best wishes. Knowledge is power. by jjwhalen

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90791 - Posted: 2 Jan 2016 | 11:57:28 UTC

A new version of "geneferocl4" is available (32-bit binary for windows) at https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/windows (for benchmarking).

It is faster than the previous version but the counterpoint is that bMax is one-half OCL3 limit (and about twice OCL limit).

I'm working on my laptop and this version is not tested. On a GeForce 740M, OCL4 is now about as fast as OCL.

Yves

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 90792 - Posted: 2 Jan 2016 | 12:06:50 UTC - in response to Message 90791.

A new version of "geneferocl4" is available (32-bit binary for windows) at https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/windows (for benchmarking).

It is faster than the previous version but the counterpoint is that bMax is one-half OCL3 limit (and about twice OCL limit).

I'm working on my laptop and this version is not tested. On a GeForce 740M, OCL4 is now about as fast as OCL.

Yves


Is it worthwhile having both the "old OCL4" and "new OCL4" transforms available? I'm envisioning the combined OCL app that can automatically switch to the best transform.
____________
My lucky number is 75898524288+1

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 90793 - Posted: 2 Jan 2016 | 13:03:05 UTC - in response to Message 90791.
Last modified: 2 Jan 2016 | 13:08:08 UTC

It is faster than the previous version but the counterpoint is that bMax is one-half OCL3 limit (and about twice OCL limit).


GTX 580:

Genefer mark scores:

Old OCL4: 48
New OCL4: 55
OCL: 77

Slightly better than the old OCL4, but nowhere close to OCL. That isn't the whole story, however. As with the first OCL4, there's a significant variance in speed depending on N:

Speed improvement of new OCL4 vs. old OCL4 and new OCL4 vs. OCL:

N v old OCL4 v OCL 32768 13.25% -27.81% 65536 0.00% -35.40% 131072 14.17% -29.17% 262144 15.25% -25.05% 524288 13.41% -30.64% 1048576 16.57% -26.86% 2097152 14.91% -30.62% 4194304 17.55% -26.67%


On average, the new OCL4 is showing about a 15% increase in speed, with the very noticeable exception of n=16, where it's benchmarking at exactly the same speed as the old OCL4.

If you recall, n=16 is also where the old OCL4 had the most improvement over OCL3. For whatever reason, the old OCL4 is particularly good at n=16 while the new OCL4 is particularly bad at n=16.

Since I'm crunching n=16 right now, the old OCL4 is the better choice since it offers the same speed but a higher b limit.
____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90794 - Posted: 2 Jan 2016 | 13:07:50 UTC - in response to Message 90792.

Is it worthwhile having both the "old OCL4" and "new OCL4" transforms available? I'm envisioning the combined OCL app that can automatically switch to the best transform.

Yes, the final version will be a single NTT that will select the residue numeral system, depending on b (and hardware?).
Then it will replace ocl3. ocl2 should run slower than a 3-prime NTT and become obsolete.
The number of RNS should not be too large because all of them should be validated (some bugs may occur with a probability ~ 1/2^64 and be difficult to find!).
Today, OCL4 are some test applications, to collect some benchmarking.

It would be possible to use a large set of RNS but I would not be confident in the result if a number is double-checked with the same RNS.
A great improvement from the point of view of the server would be to double-check a number with a different RNS (we may consider the FP-transform as a RNS).
Then double-check would detect hardware and software errors.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90796 - Posted: 2 Jan 2016 | 13:28:12 UTC - in response to Message 90791.
Last modified: 2 Jan 2016 | 14:24:11 UTC

A new version of "geneferocl4" is available (32-bit binary for windows) at https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/windows (for benchmarking).

It is faster than the previous version but the counterpoint is that bMax is one-half OCL3 limit (and about twice OCL limit).

I'm working on my laptop and this version is not tested. On a GeForce 740M, OCL4 is now about as fast as OCL.

Yves

Yo check this out: On a GTX 970, OCL4 (NEW) IS EVEN FASTER THAN OCL3!!!!!. That's insane. And if you thought OCL3 was bad on heat + power consumption, the new OCL4 is even worse: n=22 was consuming ~87% power on my card, whereas OCL4_new was pulling ~92% at the same voltage and clocks!

Can we, like, change it from "OCL4 (New)" to "OCL5 Compute Power Master Race"?

https://docs.google.com/spreadsheets/d/1gZSqRf2brAir3ornqqPP_8kMa4IPY9upfF9Cynw4p2o/edit?usp=sharing


EDIT: Added benchmarks for the AMD A6-3500 iGPU. As per my previous test, OCL 4 (Old) is significantly faster than OCL3. OCL4 (New) is even faster though, by around 20%. Congrats!

eXaPower
Send message
Joined: 30 Sep 13
Posts: 122
ID: 259902
Credit: 1,461,890,363
RAC: 1,328
Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Amethyst: Earned 1,000,000 credits (1,031,618)Cullen LLR Gold: Earned 500,000 credits (620,013)ESP LLR Amethyst: Earned 1,000,000 credits (1,043,069)Generalized Cullen/Woodall LLR Gold: Earned 500,000 credits (585,120)PPS LLR Turquoise: Earned 5,000,000 credits (9,238,059)PSP LLR Gold: Earned 500,000 credits (506,876)SoB LLR Silver: Earned 100,000 credits (379,643)SR5 LLR Turquoise: Earned 5,000,000 credits (5,022,053)SGS LLR Amethyst: Earned 1,000,000 credits (1,006,021)TRP LLR Ruby: Earned 2,000,000 credits (2,403,013)Woodall LLR Gold: Earned 500,000 credits (509,918)321 Sieve Ruby: Earned 2,000,000 credits (2,035,186)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,122,369)PPS Sieve Double Gold: Earned 500,000,000 credits (758,913,230)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,058,245)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (3,185,685)AP 26/27 Turquoise: Earned 5,000,000 credits (9,901,307)GFN Double Gold: Earned 500,000,000 credits (656,149,179)
Message 90798 - Posted: 2 Jan 2016 | 14:10:10 UTC

New OCL 4:

GT 630 (CC 3.5) @ 900MHz on PCIe 2.0 x1:


Generalized Fermat Number Bench
2199064^8192+1 Time: 240 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 261 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 383 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 736 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 1.73 ms/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 3.22 ms/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 6.29 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 12.1 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 24.3 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 47.7 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 117 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 8.


GTX 750 (CC 5.0) @ 1.4GHz on PCIe 3.0 x4:

Generalized Fermat Number Bench
2199064^8192+1 Time: 67.2 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 73.8 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 99.3 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 162 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 305 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 590 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 1.15 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 2.29 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 4.61 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 9.32 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 20.2 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 44.


GTX 970 (CC 5.2) @ 1.5GHz on PCIe 3.0 x4:

Generalized Fermat Number Bench
2199064^8192+1 Time: 84.4 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 86.1 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 88.7 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 104 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 184 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 290 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 530 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 994 us/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 1.88 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 3.63 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 8.72 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 105.


GTX 970 (CC 5.2) @ 1.5GHz on PCIe 3.0 x8 :

Generalized Fermat Number Bench
2199064^8192+1 Time: 62.1 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 58.8 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 66.1 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 84.9 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 151 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 259 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 495 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 956 us/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 1.81 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 3.61 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 8.53 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 108.




Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 90799 - Posted: 2 Jan 2016 | 15:22:38 UTC - in response to Message 90791.

A new version of "geneferocl4" is available (32-bit binary for windows) at https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/windows (for benchmarking).

It is faster than the previous version but the counterpoint is that bMax is one-half OCL3 limit (and about twice OCL limit).

I'm working on my laptop and this version is not tested. On a GeForce 740M, OCL4 is now about as fast as OCL.

Yves


The same app for Linux32 and Linux64 is available at https://www.assembla.com/spaces/genefer/subversion/commits/979
____________
Best wishes. Knowledge is power. by jjwhalen

Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 90800 - Posted: 2 Jan 2016 | 15:41:15 UTC - in response to Message 90799.

The old or the new app makes no difference for my GTS450eco (device 'GeForce GTS 450', version 'OpenCL 1.1 CUDA' and driver '352.63').

geneferocl4_linux32

2199064^8192+1 Time: 87.4 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 109 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 207 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 430 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 923 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.82 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 3.78 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 7.77 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 15.9 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 32.5 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 73.4 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 12.


geneferocl4_linux64

2199064^8192+1 Time: 67.9 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 114 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 220 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 464 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 970 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.93 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 4 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 8.05 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 16.8 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 34.6 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 75.5 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 12.

____________
Best wishes. Knowledge is power. by jjwhalen

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 90802 - Posted: 2 Jan 2016 | 16:51:42 UTC

to have some AMD GPU benchmark, Fury Nano

OCL: 131
OCL3: 70
OCL4: 86
OCL4 new build 978: 107


Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'.

2199064^8192+1 Time: 75.1 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 71.6 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 73.7 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 83.6 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 115 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 220 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 481 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 961 us/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 1.92 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 3.76 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 7.9 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 107.
Priority change succeeded.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90803 - Posted: 2 Jan 2016 | 17:00:32 UTC - in response to Message 90796.

...

Yo check this out: On a GTX 970, OCL4 (NEW) IS EVEN FASTER THAN OCL3!!!!!. That's insane. And if you thought OCL3 was bad on heat + power consumption, the new OCL4 is even worse: n=22 was consuming ~87% power on my card, whereas OCL4_new was pulling ~92% at the same voltage and clocks!

Can we, like, change it from "OCL4 (New)" to "OCL5 Compute Power Master Race"?

...


Indeed, quite an achievement, and I thought OCL3 was killer! On my 980ti@1.35GHz:
OCL3: 132
OCL4 (old): 118
OCL4 (new): 160

This leads to an important question considering the upcoming challenge: Have residues and BOINC-related things been implemented in the code? With runtimes 20% lower than OCL3, I'd really like to use this. I can run ahead of time app_info tests of course, but this will drop WR times to less than 2 days! (~44 hrs at "stock" 1.35, but I'd lower it a smidge for stability).
____________
Eating more cheese on Thursdays.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 90804 - Posted: 2 Jan 2016 | 17:18:01 UTC

Yves,

Considering the differences between "old OCL4" and "new OCL4", and given that the old OCL4 is in at least one actual use case faster than the new OCL4, could we split the new and old algorithms apart so that we can tell one from the other? Make the new one OCL5, perhaps? It would also be useful to change how they're identified in the stdout heading.

(In actual tests of numbers (not benchmarks) near the n=16 leading edge, the old OCL4 takes about 3:00 while the new one takes about 3:10.)

For sure, the n=16 behavior is an outlier, but it's one of the ranges where we'd actually be using OCL4 if it was live.

Anyway, the bottom line is that I think the new and old OCL4 transforms behave very differently under different usage scenarios and shouldn't be lumped together as a single transform.


____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90805 - Posted: 2 Jan 2016 | 17:22:25 UTC - in response to Message 90803.

OCL4 (new): 160

Great, a new record, faster than AMD 390X / OCL 139 and Titan Black / OCL 130!

This leads to an important question considering the upcoming challenge: Have residues and BOINC-related things been implemented in the code?

Yes, but I didn't run a full test (n >= 18 are not tested).

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90806 - Posted: 2 Jan 2016 | 17:22:57 UTC - in response to Message 90803.

Any chance we'll see a mac version any time soon? Or can I just grab the source, libraries, and makefile and give it a go?

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90807 - Posted: 2 Jan 2016 | 17:32:27 UTC - in response to Message 90805.

OCL4 (new): 160

Great, a new record, faster than AMD 390X / OCL 139 and Titan Black / OCL 130!

This leads to an important question considering the upcoming challenge: Have residues and BOINC-related things been implemented in the code?

Yes, but I didn't run a full test (n >= 18 are not tested).


Awesome. Once my TRPs finish I'll set up an app_info and run one of each to see what happens.
____________
Eating more cheese on Thursdays.

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 90808 - Posted: 2 Jan 2016 | 17:34:59 UTC - in response to Message 90805.

Yes, but I didn't run a full test (n >= 18 are not tested).

I'm running couple of test for known n=16 primes. So far so good...
Should I add some n=18 primes as well?

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90809 - Posted: 2 Jan 2016 | 18:17:10 UTC - in response to Message 90804.

Considering the differences between "old OCL4" and "new OCL4", and given that the old OCL4 is in at least one actual use case faster than the new OCL4, could we split the new and old algorithms apart so that we can tell one from the other?

Until now, there is a single use case, n=16 on a GTX 580 (the most important for you, I agree :o) ).
For n=16, actual b is 5,252,322 and new OCL4 can't test b > 5,720,809.
For this reason, now my goal is to extend OCL4 in the range [5,720,810 - ~11,500,000] with a code as fast as OCL3 on Maxwell and "old OCL4" on other GPU.
Then the "old OCL4" is soon going to be useless (and OCL3 too).

It would also be useful to change how they're identified in the stdout heading.

Old header is:
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2/32-bit)
Copyright 2001-2015, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2015, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

New one is:
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.


NeoMetal*
Volunteer tester
Avatar
Send message
Joined: 25 Mar 11
Posts: 418
ID: 92179
Credit: 1,747,428,303
RAC: 0
Eliminated 1 conjecture "k"Found 1 prime in the 2018 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (7,698,294)Cullen LLR Turquoise: Earned 5,000,000 credits (6,655,866)ESP LLR Turquoise: Earned 5,000,000 credits (8,198,062)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,817,032)PPS LLR Jade: Earned 10,000,000 credits (10,016,096)PSP LLR Jade: Earned 10,000,000 credits (10,891,779)SoB LLR Jade: Earned 10,000,000 credits (11,111,741)SR5 LLR Turquoise: Earned 5,000,000 credits (7,278,494)SGS LLR Turquoise: Earned 5,000,000 credits (7,457,856)TRP LLR Turquoise: Earned 5,000,000 credits (7,714,186)Woodall LLR Turquoise: Earned 5,000,000 credits (5,726,778)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,094,130)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (11,628,594)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,040,865,445)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,235,150)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,376,844)AP 26/27 Jade: Earned 10,000,000 credits (11,615,539)GFN Double Silver: Earned 200,000,000 credits (427,593,017)PSA Double Bronze: Earned 100,000,000 credits (125,462,363)
Message 90811 - Posted: 2 Jan 2016 | 20:38:53 UTC

Yves, if you are not changing the new OCL4 to OCL5 maybe change it to OCL4.1
____________
Largest Primes to Date:
As Double Checker: SR5 109208*5^1816285+1 Dgts-1,269,534
As Initial Finder: SR5 243944*5^1258576-1 Dgts-879,713


Iain BethuneProject donor
Honorary cruncher
Send message
Joined: 28 Jan 09
Posts: 1588
ID: 34775
Credit: 194,056,043
RAC: 0
321 LLR Gold: Earned 500,000 credits (597,175)Cullen LLR Amethyst: Earned 1,000,000 credits (1,355,381)ESP LLR Turquoise: Earned 5,000,000 credits (5,048,594)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,564,412)PPS LLR Amethyst: Earned 1,000,000 credits (1,025,115)PSP LLR Sapphire: Earned 20,000,000 credits (28,367,849)SoB LLR Jade: Earned 10,000,000 credits (16,215,329)SR5 LLR Turquoise: Earned 5,000,000 credits (5,189,992)SGS LLR Amethyst: Earned 1,000,000 credits (1,244,067)TRP LLR Turquoise: Earned 5,000,000 credits (5,169,405)Woodall LLR Amethyst: Earned 1,000,000 credits (1,070,956)321 Sieve Bronze: Earned 10,000 credits (20,003)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (200,371)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (11,645,025)PPS Sieve Turquoise: Earned 5,000,000 credits (7,536,532)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,017,144)TRP Sieve (suspended) Gold: Earned 500,000 credits (554,946)AP 26/27 Ruby: Earned 2,000,000 credits (3,577,848)GFN Emerald: Earned 50,000,000 credits (97,047,976)PSA Ruby: Earned 2,000,000 credits (4,606,694)
Message 90814 - Posted: 2 Jan 2016 | 21:31:05 UTC - in response to Message 90806.
Last modified: 2 Jan 2016 | 22:16:54 UTC

Any chance we'll see a mac version any time soon? Or can I just grab the source, libraries, and makefile and give it a go?


I was busy working on the combined OCL app, so I hadn't played with OCL4 yet, but I have just built an app which you can find in https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/mac.

Use it with caution! On my card at least, there may be a bug the transforms for n=14 and some others appear to show a round-off error in the benchmarks. However, if I run those candidates manually, I see a correct result, although OCL is faster than OCL4, contradicting the results of the benchmarks…

Using OCL transform Running on platform 'Apple', device 'ATI Radeon HD - FirePro D700 Compute Engine', version 'OpenCL 1.2 ' and driver '1.2 (Oct 13 2015 18:31:41)'. ... 1798620^16384+1 is composite. (RES=0e881a1bcf7a1253) (102481 digits) (err = 0.2324) (time = 0:00:41) 21:35:38 Using OCL4 transform Running on platform 'Apple', device 'ATI Radeon HD - FirePro D700 Compute Engine', version 'OpenCL 1.2 ' and driver '1.2 (Oct 13 2015 18:31:41)'. 1798620^16384+1 is composite. (RES=0e881a1bcf7a1253) (102481 digits) (err = 0.0000) (time = 0:01:10) 21:34:54


For reference, benchmarks on my ATI FirePro D700. Looks like it is comfortably faster than OCL3.

Running benchmarks for transform implementation "OCL" 2199064^8192+1 Time: 119 us/mul. Err: 0.1953 51956 digits 1798620^16384+1 Time: 118 us/mul. Err: 0.2009 102481 digits 1471094^32768+1 Time: 124 us/mul. Err: 0.2246 202102 digits 1203210^65536+1 Time: 145 us/mul. Err: 0.1875 398482 digits 984108^131072+1 Time: 197 us/mul. Err: 0.1953 785521 digits 804904^262144+1 Time: 367 us/mul. Err: 0.1917 1548156 digits 658332^524288+1 Time: 652 us/mul. Err: 0.1968 3050541 digits 538452^1048576+1 Time: 1.38 ms/mul. Err: 0.1914 6009544 digits 440400^2097152+1 Time: 2.61 ms/mul. Err: 0.1912 11836006 digits 360204^4194304+1 Time: 5.08 ms/mul. Err: 0.1875 23305854 digits 294612^8388608+1 Time: 10.3 ms/mul. Err: 0.1855 45879398 digits Genefer Mark = 79. Running benchmarks for transform implementation "OCL3" 2199064^8192+1 Time: 277 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 256 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 278 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 314 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 467 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 760 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 1.3 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 2.5 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 4.68 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 9.35 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 18.3 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 44. Running benchmarks for transform implementation "OCL4" 2199064^8192+1 Time: 185 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 196 us/mul. Err: 1.0000 102481 digits 1471094^32768+1 Time: 212 us/mul. Err: 1.0000 202102 digits 1203210^65536+1 Time: 241 us/mul. Err: 1.0000 398482 digits 984108^131072+1 Time: 353 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 595 us/mul. Err: 1.0000 1548156 digits 658332^524288+1 Time: 1.07 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 1.98 ms/mul. Err: 1.0000 6009544 digits 440400^2097152+1 Time: 3.68 ms/mul. Err: 1.0000 11836006 digits 360204^4194304+1 Time: 7.21 ms/mul. Err: 1.0000 23305854 digits 294612^8388608+1 Time: 14 ms/mul. Err: 1.0000 45879398 digits Genefer Mark = 56.


Cheers

- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime!

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90815 - Posted: 2 Jan 2016 | 21:48:33 UTC - in response to Message 90814.

Any chance we'll see a mac version any time soon? Or can I just grab the source, libraries, and makefile and give it a go?


I was busy working on the combined OCL app, so I hadn't played with OCL4 yet, but I have just built an app which you can find in https://www.assembla.com/spaces/genefer/subversion/source/HEAD/trunk/bin/mac.

Use it with caution! On my card at least, there may be a bug the transforms for n=14 and some others appear to show a round-off error in the benchmarks. However, if I run those candidates manually, I see a correct result, although OCL is faster than OCL4, contradicting the results of the benchmarks…


Excellent! I'll give it a try, I dropped a 980 into an old mac pro, and would love to see how it does.

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90816 - Posted: 2 Jan 2016 | 21:55:03 UTC

Mac pro 5,1 w/980

Benchmarks for the ocl executables available
ocl 71
ocl2 17
ocl3 86
ocl4 105

Holy cow!

ReaDy
Send message
Joined: 9 Aug 13
Posts: 3
ID: 247721
Credit: 33,380,984
RAC: 306
321 LLR Silver: Earned 100,000 credits (179,228)Cullen LLR Silver: Earned 100,000 credits (102,307)ESP LLR Silver: Earned 100,000 credits (129,306)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (106,193)PPS LLR Ruby: Earned 2,000,000 credits (2,078,827)PSP LLR Silver: Earned 100,000 credits (115,249)SoB LLR Gold: Earned 500,000 credits (610,618)SR5 LLR Gold: Earned 500,000 credits (524,530)SGS LLR Gold: Earned 500,000 credits (906,356)TRP LLR Silver: Earned 100,000 credits (123,964)Woodall LLR Silver: Earned 100,000 credits (107,692)Generalized Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (503,531)PPS Sieve Turquoise: Earned 5,000,000 credits (5,363,261)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Gold: Earned 500,000 credits (500,141)TRP Sieve (suspended) Gold: Earned 500,000 credits (501,953)AP 26/27 Ruby: Earned 2,000,000 credits (2,106,403)GFN Jade: Earned 10,000,000 credits (12,833,687)PSA Turquoise: Earned 5,000,000 credits (6,075,737)
Message 90817 - Posted: 2 Jan 2016 | 21:59:22 UTC

GTX 750 Ti @ 1.2GHz on PCIe 3.0 x8:
OCL: 28
OCL2: 14
OCL3: 33
OCL4 (rev 978): 46

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 750 Ti', version 'OpenCL 1.2 CUDA' and driver '359.06' Running benchmarks for transform implementation "OCL4" 2199064^8192+1 Time: 49.4 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 56.1 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 82.6 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 132 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 270 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 532 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 1.07 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 2.15 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 4.36 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 8.91 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 19.8 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 46.

AMD HD 7790 @ 1GHz on PCIe 3.0 x8(no primegrid crunch, only test):
OCL: 31
OCL2: fail (maybe driver problem)
OCL3: 14
OCL4 (rev 978): 26

Running on platform 'AMD Accelerated Parallel Processing', device 'Bonaire', version 'OpenCL 1.2 AMD-APP (1411.4)' and driver '1411.4 (VM)' Running benchmarks for transform implementation "OCL4" 2199064^8192+1 Time: 98.7 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 164 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 142 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 229 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 493 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 970 us/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 1.9 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 3.71 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 7.46 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 15.2 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 35.2 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 26.

And benchmarks OCL4 (rev 978) for the Intel HD 4000 ( i7 3770 iGPU)

Running on platform 'Intel(R) OpenCL', device 'Intel(R) HD Graphics 4000', version 'OpenCL 1.2 ' and driver '10.18.10.3621' Running benchmarks for transform implementation "OCL4" 2199064^8192+1 Time: 201 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 334 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 603 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 1.18 ms/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 2.35 ms/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 4.72 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 9.68 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 19.8 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 41.9 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 87.3 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 177 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 5.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90818 - Posted: 2 Jan 2016 | 22:21:15 UTC

I found a little problem when trying to run the app. I decided to run it on my laptop with Intel HD; I didn't expect it to have good performance, I just wanted to see if the program worked..... and it didn't.

When trying to open the exe, the console window just freezes. You press enter and the app never opens, the cursor just stays there. In fact, you can't even use the Ctr+C command to stop it, it just stays stuck for some reason.

With OCL 1~3, I have no problems openning the app, but OCL4 is failing me.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90821 - Posted: 2 Jan 2016 | 23:04:59 UTC - in response to Message 90818.

I found a little problem when trying to run the app. I decided to run it on my laptop with Intel HD; I didn't expect it to have good performance, I just wanted to see if the program worked..... and it didn't.

When trying to open the exe, the console window just freezes. You press enter and the app never opens, the cursor just stays there. In fact, you can't even use the Ctr+C command to stop it, it just stays stuck for some reason.

With OCL 1~3, I have no problems openning the app, but OCL4 is failing me.

It runs on my 'Intel(R) HD Graphics 4000' (Windows 10, driver '10.18.10.4276').
What's your OS?

I compiled ocl4 with Visual Studio 2015, previous versions were compiled with VS 2013...

Profile TylerProject donor
Project administrator
Volunteer tester
Send message
Joined: 4 Dec 12
Posts: 1077
ID: 183129
Credit: 1,280,170,555
RAC: 7,054
Discovered 1 mega primeFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (5,238,816)Cullen LLR Turquoise: Earned 5,000,000 credits (5,124,261)ESP LLR Turquoise: Earned 5,000,000 credits (8,479,525)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,656,837)PPS LLR Sapphire: Earned 20,000,000 credits (23,667,318)PSP LLR Sapphire: Earned 20,000,000 credits (37,386,303)SoB LLR Sapphire: Earned 20,000,000 credits (21,004,651)SR5 LLR Sapphire: Earned 20,000,000 credits (33,215,655)SGS LLR Turquoise: Earned 5,000,000 credits (5,000,408)TRP LLR Jade: Earned 10,000,000 credits (12,199,250)Woodall LLR Turquoise: Earned 5,000,000 credits (8,449,472)321 Sieve Turquoise: Earned 5,000,000 credits (8,012,059)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (27,921,016)PPS Sieve Double Gold: Earned 500,000,000 credits (542,336,593)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (7,990,608)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,217,057)AP 26/27 Turquoise: Earned 5,000,000 credits (7,111,637)GFN Sapphire: Earned 20,000,000 credits (31,676,706)PSA Double Silver: Earned 200,000,000 credits (420,586,383)
Message 90823 - Posted: 3 Jan 2016 | 0:58:40 UTC

GTX 760 Stock:
OCL: 54
OCL2: 12
OCL3: 37
OCL4(v1): 48
OCL4(v2): 56

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90824 - Posted: 3 Jan 2016 | 1:10:32 UTC - in response to Message 90821.

I found a little problem when trying to run the app. I decided to run it on my laptop with Intel HD; I didn't expect it to have good performance, I just wanted to see if the program worked..... and it didn't.

When trying to open the exe, the console window just freezes. You press enter and the app never opens, the cursor just stays there. In fact, you can't even use the Ctr+C command to stop it, it just stays stuck for some reason.

With OCL 1~3, I have no problems openning the app, but OCL4 is failing me.

It runs on my 'Intel(R) HD Graphics 4000' (Windows 10, driver '10.18.10.4276').
What's your OS?

I compiled ocl4 with Visual Studio 2015, previous versions were compiled with VS 2013...

This is the computer. OS is Windows 10, upgraded from 8 (not 8.1).

I've got no idea on what Visual Studio is, so if there's something you'd like me download and see if it fixes the issue, just give me a link and I'll try it.

ReaDy
Send message
Joined: 9 Aug 13
Posts: 3
ID: 247721
Credit: 33,380,984
RAC: 306
321 LLR Silver: Earned 100,000 credits (179,228)Cullen LLR Silver: Earned 100,000 credits (102,307)ESP LLR Silver: Earned 100,000 credits (129,306)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (106,193)PPS LLR Ruby: Earned 2,000,000 credits (2,078,827)PSP LLR Silver: Earned 100,000 credits (115,249)SoB LLR Gold: Earned 500,000 credits (610,618)SR5 LLR Gold: Earned 500,000 credits (524,530)SGS LLR Gold: Earned 500,000 credits (906,356)TRP LLR Silver: Earned 100,000 credits (123,964)Woodall LLR Silver: Earned 100,000 credits (107,692)Generalized Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (503,531)PPS Sieve Turquoise: Earned 5,000,000 credits (5,363,261)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Gold: Earned 500,000 credits (500,141)TRP Sieve (suspended) Gold: Earned 500,000 credits (501,953)AP 26/27 Ruby: Earned 2,000,000 credits (2,106,403)GFN Jade: Earned 10,000,000 credits (12,833,687)PSA Turquoise: Earned 5,000,000 credits (6,075,737)
Message 90831 - Posted: 3 Jan 2016 | 5:57:04 UTC

I've got no idea on what Visual Studio is, so if there's something you'd like me download and see if it fixes the issue, just give me a link and I'll try it.

Maybe need install Visual C++ Redistributable for Visual Studio 2015
Your choose vc_redist.x64.exe

Kwartet!Project donor
Send message
Joined: 22 Nov 12
Posts: 21
ID: 181030
Credit: 123,029,507
RAC: 0
321 LLR Amethyst: Earned 1,000,000 credits (1,336,554)Cullen LLR Amethyst: Earned 1,000,000 credits (1,114,814)ESP LLR Amethyst: Earned 1,000,000 credits (1,004,266)Generalized Cullen/Woodall LLR Gold: Earned 500,000 credits (618,684)PPS LLR Amethyst: Earned 1,000,000 credits (1,058,191)PSP LLR Amethyst: Earned 1,000,000 credits (1,367,424)SoB LLR Amethyst: Earned 1,000,000 credits (1,691,911)SR5 LLR Amethyst: Earned 1,000,000 credits (1,226,031)SGS LLR Amethyst: Earned 1,000,000 credits (1,236,202)TRP LLR Amethyst: Earned 1,000,000 credits (1,097,559)Woodall LLR Amethyst: Earned 1,000,000 credits (1,170,522)Generalized Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (2,001,505)PPS Sieve Emerald: Earned 50,000,000 credits (53,184,267)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,007,577)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,999,175)AP 26/27 Bronze: Earned 10,000 credits (20,215)GFN Sapphire: Earned 20,000,000 credits (49,288,593)PSA Gold: Earned 500,000 credits (606,016)
Message 90836 - Posted: 3 Jan 2016 | 9:47:55 UTC

HD7950 Boost @ 1100/1400 MHz
OCL: 121
OCL2: 12
OCL3: 51
OCL4: 72

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90838 - Posted: 3 Jan 2016 | 10:38:16 UTC - in response to Message 90816.
Last modified: 3 Jan 2016 | 10:42:16 UTC

Mac pro 5,1 w/980

Benchmarks for the ocl executables available
ocl 71
ocl2 17
ocl3 86
ocl4 105

Holy cow!



WR unit running for 12.5 hours, looking at a 72-73 hour completion. Straight up ocl was taking 113 hours. (based on time running and % complete)

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90839 - Posted: 3 Jan 2016 | 10:41:52 UTC - in response to Message 90831.

I've got no idea on what Visual Studio is, so if there's something you'd like me download and see if it fixes the issue, just give me a link and I'll try it.

Maybe need install Visual C++ Redistributable for Visual Studio 2015
Your choose vc_redist.x64.exe

I installed it, but the app still won't run.

Profile DeleteNullProject donor
Volunteer tester
Avatar
Send message
Joined: 6 Apr 06
Posts: 242
ID: 2663
Credit: 7,255,283,251
RAC: 126,806
Discovered 26 mega primesEliminated 3 conjecture "k"sFound 4 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 2 mega primes in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 2 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Emerald: Earned 50,000,000 credits (60,657,554)Cullen LLR Emerald: Earned 50,000,000 credits (67,814,375)ESP LLR Emerald: Earned 50,000,000 credits (61,429,176)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (59,199,415)PPS LLR Double Silver: Earned 200,000,000 credits (246,743,264)PSP LLR Emerald: Earned 50,000,000 credits (76,449,391)SoB LLR Emerald: Earned 50,000,000 credits (71,933,705)SR5 LLR Double Bronze: Earned 100,000,000 credits (102,395,858)SGS LLR Emerald: Earned 50,000,000 credits (54,295,707)TPS LLR (retired) Bronze: Earned 10,000 credits (61,785)TRP LLR Double Bronze: Earned 100,000,000 credits (102,088,428)Woodall LLR Emerald: Earned 50,000,000 credits (50,544,668)321 Sieve Emerald: Earned 50,000,000 credits (63,039,849)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,685,309)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (53,324,439)PPS Sieve Double Ruby: Earned 2,000,000,000 credits (3,787,521,227)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,057,785)TRP Sieve (suspended) Emerald: Earned 50,000,000 credits (50,146,560)AP 26/27 Double Silver: Earned 200,000,000 credits (254,699,080)GFN Double Amethyst: Earned 1,000,000,000 credits (1,799,551,280)PSA Double Silver: Earned 200,000,000 credits (204,786,115)
Message 90840 - Posted: 3 Jan 2016 | 11:30:53 UTC

R9 390X:
OCL4 (old): 90
OCL4 (new): 101

____________
DeleteNull

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90841 - Posted: 3 Jan 2016 | 11:55:31 UTC - in response to Message 90839.

I installed it, but the app still won't run.

If ocl3 runs and not ocl4, I don't understand.
Both depend on a single library: C:\windows\system32\OpenCL.dll
When they start, they just print a message to the console...
Anti-virus program or Internet security?

mackerelProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Oct 08
Posts: 2460
ID: 29980
Credit: 442,802,854
RAC: 10,291
Discovered 4 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 5 primes in the 2019 Tour de PrimesFound 6 primes in the 2020 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (8,774,878)Cullen LLR Turquoise: Earned 5,000,000 credits (5,903,451)ESP LLR Turquoise: Earned 5,000,000 credits (6,454,573)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,122,074)PPS LLR Emerald: Earned 50,000,000 credits (83,377,417)PSP LLR Jade: Earned 10,000,000 credits (15,223,714)SoB LLR Jade: Earned 10,000,000 credits (17,319,914)SR5 LLR Sapphire: Earned 20,000,000 credits (23,996,561)SGS LLR Turquoise: Earned 5,000,000 credits (7,342,780)TPS LLR (retired) Bronze: Earned 10,000 credits (34,130)TRP LLR Jade: Earned 10,000,000 credits (19,866,589)Woodall LLR Turquoise: Earned 5,000,000 credits (8,171,820)321 Sieve Sapphire: Earned 20,000,000 credits (20,236,219)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,383,853)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,626,419)PPS Sieve Emerald: Earned 50,000,000 credits (76,969,144)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,293,882)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,012,757)AP 26/27 Sapphire: Earned 20,000,000 credits (21,918,894)GFN Emerald: Earned 50,000,000 credits (76,466,089)PSA Ruby: Earned 2,000,000 credits (2,939,755)
Message 90843 - Posted: 3 Jan 2016 | 12:27:08 UTC

GTX 980 Ti

OCL3: 120
OCL4 (old): 105
OCL4 (new): 154 (+47% over old OCL4)


R9 280X

OCL3: 55
OCL4 (old): 60
OCL4 (new): 72 (+20% over old OCL4)

ReaDy
Send message
Joined: 9 Aug 13
Posts: 3
ID: 247721
Credit: 33,380,984
RAC: 306
321 LLR Silver: Earned 100,000 credits (179,228)Cullen LLR Silver: Earned 100,000 credits (102,307)ESP LLR Silver: Earned 100,000 credits (129,306)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (106,193)PPS LLR Ruby: Earned 2,000,000 credits (2,078,827)PSP LLR Silver: Earned 100,000 credits (115,249)SoB LLR Gold: Earned 500,000 credits (610,618)SR5 LLR Gold: Earned 500,000 credits (524,530)SGS LLR Gold: Earned 500,000 credits (906,356)TRP LLR Silver: Earned 100,000 credits (123,964)Woodall LLR Silver: Earned 100,000 credits (107,692)Generalized Cullen/Woodall Sieve (suspended) Gold: Earned 500,000 credits (503,531)PPS Sieve Turquoise: Earned 5,000,000 credits (5,363,261)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Gold: Earned 500,000 credits (500,141)TRP Sieve (suspended) Gold: Earned 500,000 credits (501,953)AP 26/27 Ruby: Earned 2,000,000 credits (2,106,403)GFN Jade: Earned 10,000,000 credits (12,833,687)PSA Turquoise: Earned 5,000,000 credits (6,075,737)
Message 90844 - Posted: 3 Jan 2016 | 12:54:03 UTC

I with use of app_info.xml started OCL4 (rev 978) in tasks GFN-17-Low and GFN-18.
Part of tasks passed confirmation, the part waits for validation.
Except reduction of time of calculations loading of cpu considerably decreased!

i7 3770 HT on + GTX 750 Ti @ 1.2GHz on PCIe 3.0 x8:

GFN-17-Low OCL3 (default apps) - cpu used 12...13% | time task 18min
GFN-17-Low OCL4 (rev 978) - cpu used 01...04% | time task 13min

GFN-18 OCL3 (default apps) - cpu used 12...13% | time task 1:10 hour
GFN-18 OCL4 (rev 978) - cpu used 00...01% | time task 50min

It is very good.

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 90849 - Posted: 3 Jan 2016 | 14:47:17 UTC
Last modified: 3 Jan 2016 | 14:48:11 UTC

R9 290x
ocl 108
ocl3 69
ocl4 81

Don't forget to update those copyright notices...

:)

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90850 - Posted: 3 Jan 2016 | 15:27:37 UTC - in response to Message 90844.

Except reduction of time of calculations loading of cpu considerably decreased!

I forget to notice this optimisation.
I removed GPU synchronisation on NVIDIA (and let the driver do it). With the latest versions of the driver, CPU load is a few percent, it is no longer one core.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90853 - Posted: 3 Jan 2016 | 17:16:55 UTC - in response to Message 90850.

Except reduction of time of calculations loading of cpu considerably decreased!

I forget to notice this optimisation.
I removed GPU synchronisation on NVIDIA (and let the driver do it). With the latest versions of the driver, CPU load is a few percent, it is no longer one core.



Huzzah! 99%GPU, 0% CPU (361.43). That should be changed on the other versions as well :-)

Currently BOINC testing a GFN-21 task on my 980ti, estimated time ~11hrs at reduced clocks for stability.
____________
Eating more cheese on Thursdays.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 90859 - Posted: 3 Jan 2016 | 21:21:27 UTC - in response to Message 90850.

I forget to notice this optimisation.
I removed GPU synchronisation on NVIDIA (and let the driver do it). With the latest versions of the driver, CPU load is a few percent, it is no longer one core.


I can confirm that the new OCL4 uses almost no CPU. This is with driver 359.06. Thanks! (To both you and Nvidia.)

____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90863 - Posted: 4 Jan 2016 | 3:19:22 UTC - in response to Message 90841.

I installed it, but the app still won't run.

If ocl3 runs and not ocl4, I don't understand.
Both depend on a single library: C:\windows\system32\OpenCL.dll
When they start, they just print a message to the console...
Anti-virus program or Internet security?

What happens is this: when I enter the "geneferocl4_windows" command in the console windows and press enter, it just goes to the next line and stop, without printing anything else. The cursor just stays blinking (as if it was loading), but nothing ever happens.

When running the app directly, mouse just shows that "loading in second plane" icon. Again, nothing happens. There isn't even a popup asking to allow the app to run (like the others), just plain nothingness.

Profile Wingless WonderProject donor
Avatar
Send message
Joined: 25 Dec 12
Posts: 174
ID: 186381
Credit: 394,309,812
RAC: 2,620
Discovered 1 mega primeFound 3 primes in the 2018 Tour de Primes321 LLR Amethyst: Earned 1,000,000 credits (1,374,650)Cullen LLR Amethyst: Earned 1,000,000 credits (1,661,696)ESP LLR Gold: Earned 500,000 credits (710,747)Generalized Cullen/Woodall LLR Amethyst: Earned 1,000,000 credits (1,205,397)PPS LLR Ruby: Earned 2,000,000 credits (2,397,200)PSP LLR Ruby: Earned 2,000,000 credits (2,113,325)SoB LLR Ruby: Earned 2,000,000 credits (2,921,805)SR5 LLR Amethyst: Earned 1,000,000 credits (1,284,829)SGS LLR Gold: Earned 500,000 credits (632,629)TRP LLR Turquoise: Earned 5,000,000 credits (5,321,051)Woodall LLR Silver: Earned 100,000 credits (127,395)321 Sieve Bronze: Earned 10,000 credits (88,433)Generalized Cullen/Woodall Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,626,395)PPS Sieve Double Bronze: Earned 100,000,000 credits (135,099,567)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Gold: Earned 500,000 credits (555,042)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,089,296)AP 26/27 Turquoise: Earned 5,000,000 credits (6,347,510)GFN Double Silver: Earned 200,000,000 credits (228,752,844)
Message 90865 - Posted: 4 Jan 2016 | 3:54:14 UTC - in response to Message 90863.
Last modified: 4 Jan 2016 | 3:56:46 UTC

What happens is this: when I enter the "geneferocl4_windows" command in the console windows and press enter, it just goes to the next line and stop, without printing anything else. The cursor just stays blinking (as if it was loading), but nothing ever happens.

When running the app directly, mouse just shows that "loading in second plane" icon. Again, nothing happens. There isn't even a popup asking to allow the app to run (like the others), just plain nothingness.

If you're trying to run the OCL4 benchmark, you must use the appropriate command line switch. For genererocl4_windows.exe, I think the command line should be:

geneferocl4_windows -b

EDIT: the above may run the Intel graphics. For NVIDIA, use:

geneferocl4_windows -d 1 -b

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 90881 - Posted: 4 Jan 2016 | 17:48:56 UTC - in response to Message 90865.
Last modified: 4 Jan 2016 | 17:51:16 UTC

What happens is this: when I enter the "geneferocl4_windows" command in the console windows and press enter, it just goes to the next line and stop, without printing anything else. The cursor just stays blinking (as if it was loading), but nothing ever happens.

When running the app directly, mouse just shows that "loading in second plane" icon. Again, nothing happens. There isn't even a popup asking to allow the app to run (like the others), just plain nothingness.

If you're trying to run the OCL4 benchmark, you must use the appropriate command line switch. For genererocl4_windows.exe, I think the command line should be:

geneferocl4_windows -b

EDIT: the above may run the Intel graphics. For NVIDIA, use:

geneferocl4_windows -d 1 -b

I was actually using the pure command without any arguments (in other words, the interactive mode). Pretty sure I tried it with arguments as well, but I'll check later once I get home.

But I'm sure the interactive mode works onmy other machines, as that's what I've used to get those benchmarks.

waffleironhead
Send message
Joined: 1 Oct 09
Posts: 31
ID: 47712
Credit: 77,419,417
RAC: 2,434
321 LLR Gold: Earned 500,000 credits (587,446)Cullen LLR Gold: Earned 500,000 credits (581,927)ESP LLR Gold: Earned 500,000 credits (504,268)Generalized Cullen/Woodall LLR Silver: Earned 100,000 credits (282,842)PPS LLR Amethyst: Earned 1,000,000 credits (1,203,726)PSP LLR Silver: Earned 100,000 credits (376,066)SoB LLR Amethyst: Earned 1,000,000 credits (1,456,054)SR5 LLR Gold: Earned 500,000 credits (506,833)SGS LLR Gold: Earned 500,000 credits (677,544)TRP LLR Gold: Earned 500,000 credits (509,346)Woodall LLR Gold: Earned 500,000 credits (636,086)321 Sieve Turquoise: Earned 5,000,000 credits (5,003,126)Cullen/Woodall Sieve (suspended) Bronze: Earned 10,000 credits (17,528)Generalized Cullen/Woodall Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,209,152)PPS Sieve Sapphire: Earned 20,000,000 credits (29,623,046)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,075,699)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,178,769)AP 26/27 Bronze: Earned 10,000 credits (90,113)GFN Jade: Earned 10,000,000 credits (17,968,527)PSA Gold: Earned 500,000 credits (743,320)
Message 90909 - Posted: 6 Jan 2016 | 5:33:54 UTC - in response to Message 90626.

AMD HD6950
ocl=43
ocl2=6
ocl3=18
ocl4=24



Running benchmarks for transform implementation "OCL" Running on platform 'AMD Accelerated Parallel Processing', device 'Cayman', vers ion 'OpenCL 1.2 AMD-APP (1800.8)' and driver '1800.8 (VM)'. 2199064^8192+1 Time: 71.7 us/mul. Err: 0.1875 51956 digits 1798620^16384+1 Time: 74.2 us/mul. Err: 0.1875 102481 digits 1471094^32768+1 Time: 89 us/mul. Err: 0.1875 202102 digits 1203210^65536+1 Time: 135 us/mul. Err: 0.1875 398482 digits 984108^131072+1 Time: 244 us/mul. Err: 0.1797 785521 digits 804904^262144+1 Time: 661 us/mul. Err: 0.1719 1548156 digits 658332^524288+1 Time: 1.3 ms/mul. Err: 0.1797 3050541 digits 538452^1048576+1 Time: 2.46 ms/mul. Err: 0.1758 6009544 digits 440400^2097152+1 Time: 4.79 ms/mul. Err: 0.1719 11836006 digits 360204^4194304+1 Time: 9.57 ms/mul. Err: 0.1621 23305854 digits 294612^8388608+1 Time: 19.2 ms/mul. Err: 0.1621 45879398 digits Genefer Mark = 43. Running benchmarks for transform implementation "OCL2" Running on platform 'AMD Accelerated Parallel Processing', device 'Cayman', vers ion 'OpenCL 1.2 AMD-APP (1800.8)' and driver '1800.8 (VM)'. 2199064^8192+1 Time: 460 us/mul. Err: 0.0001 51956 digits 1798620^16384+1 Time: 647 us/mul. Err: 0.0001 102481 digits 1471094^32768+1 Time: 778 us/mul. Err: 0.0001 202102 digits 1203210^65536+1 Time: 829 us/mul. Err: 0.0001 398482 digits 984108^131072+1 Time: 1.67 ms/mul. Err: 0.0001 785521 digits 804904^262144+1 Time: 4.3 ms/mul. Err: 0.0001 1548156 digits 658332^524288+1 Time: 14 ms/mul. Err: 0.0001 3050541 digits 538452^1048576+1 Time: 33.9 ms/mul. Err: 0.0001 6009544 digits 440400^2097152+1 Time: 56.6 ms/mul. Err: 0.0001 11836006 digits 360204^4194304+1 Time: 49.7 ms/mul. Err: 0.0001 23305854 digits 294612^8388608+1 Time: 105 ms/mul. Err: 0.0001 45879398 digits Genefer Mark = 6. Priority change succeeded. Running benchmarks for transform implementation "OCL3" Running on platform 'AMD Accelerated Parallel Processing', device 'Cayman', vers ion 'OpenCL 1.2 AMD-APP (1800.8)' and driver '1800.8 (VM)'. 2199064^8192+1 Time: 111 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 126 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 176 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 388 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 700 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.46 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 2.8 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 5.49 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 11.2 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 23.3 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 44.4 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 18. Priority change succeeded. Running benchmarks for transform implementation "OCL4" Running on platform 'AMD Accelerated Parallel Processing', device 'Cayman', vers ion 'OpenCL 1.2 AMD-APP (1800.8)' and driver '1800.8 (VM)'. 2199064^8192+1 Time: 94.8 us/mul. Err: 0.0000 51956 digits 1798620^16384+1 Time: 105 us/mul. Err: 0.0000 102481 digits 1471094^32768+1 Time: 140 us/mul. Err: 0.0000 202102 digits 1203210^65536+1 Time: 263 us/mul. Err: 0.0000 398482 digits 984108^131072+1 Time: 564 us/mul. Err: 0.0000 785521 digits 804904^262144+1 Time: 1.22 ms/mul. Err: 0.0000 1548156 digits 658332^524288+1 Time: 2.27 ms/mul. Err: 0.0000 3050541 digits 538452^1048576+1 Time: 4.29 ms/mul. Err: 0.0000 6009544 digits 440400^2097152+1 Time: 8.65 ms/mul. Err: 0.0000 11836006 digits 360204^4194304+1 Time: 17.5 ms/mul. Err: 0.0000 23305854 digits 294612^8388608+1 Time: 34.9 ms/mul. Err: 0.0000 45879398 digits Genefer Mark = 24. Priority change succeeded.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90923 - Posted: 6 Jan 2016 | 15:52:46 UTC

I'll just leave this here...

http://www.primegrid.com/workunit.php?wuid=455870759

The fact that my 980ti is underclocked to 1277 core/6610 memory and the other GPU is a Titan just astounds me in the result
____________
Eating more cheese on Thursdays.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 90927 - Posted: 6 Jan 2016 | 17:08:12 UTC - in response to Message 90923.

I'll just leave this here...

http://www.primegrid.com/workunit.php?wuid=455870759

The fact that my 980ti is underclocked to 1277 core/6610 memory and the other GPU is a Titan just astounds me in the result


20 million digits in less than two days! Very good :o)

Core clock is overclocked (and memory underclocked).
If genefer is not 'memory limited' the result is not so astounding.
Your 980ti 'processing power' is 7192 GFlops, a Titan Black is 5121.
Maxwell is more power efficient than Kepler, that's probably one of the reasons.

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 90933 - Posted: 6 Jan 2016 | 18:09:29 UTC - in response to Message 90927.

I'll just leave this here...

http://www.primegrid.com/workunit.php?wuid=455870759

The fact that my 980ti is underclocked to 1277 core/6610 memory and the other GPU is a Titan just astounds me in the result


20 million digits in less than two days! Very good :o)

Core clock is overclocked (and memory underclocked).
If genefer is not 'memory limited' the result is not so astounding.
Your 980ti 'processing power' is 7192 GFlops, a Titan Black is 5121.
Maxwell is more power efficient than Kepler, that's probably one of the reasons.


It's a pre-OC'd model, so 1277 is as low as I could get the slider to go, it normally runs near 1.4GHz as "stock". The memory controller was running at near 90% load during the run, I wonder what effect HBM2 will have with the next generation of cards.
____________
Eating more cheese on Thursdays.

stream
Volunteer moderator
Project administrator
Volunteer developer
Volunteer tester
Send message
Joined: 1 Mar 14
Posts: 834
ID: 301928
Credit: 488,466,944
RAC: 2,246
Discovered 1 mega primeFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,011,570)Cullen LLR Jade: Earned 10,000,000 credits (10,009,374)ESP LLR Jade: Earned 10,000,000 credits (10,009,221)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (10,012,217)PPS LLR Jade: Earned 10,000,000 credits (11,055,307)PSP LLR Jade: Earned 10,000,000 credits (10,044,081)SoB LLR Jade: Earned 10,000,000 credits (10,064,750)SR5 LLR Jade: Earned 10,000,000 credits (10,002,051)SGS LLR Jade: Earned 10,000,000 credits (10,001,215)TRP LLR Jade: Earned 10,000,000 credits (10,002,411)Woodall LLR Jade: Earned 10,000,000 credits (10,013,921)321 Sieve Sapphire: Earned 20,000,000 credits (20,004,228)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,047,667)PPS Sieve Sapphire: Earned 20,000,000 credits (20,866,490)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,043,271)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,015,177)AP 26/27 Sapphire: Earned 20,000,000 credits (20,045,194)GFN Emerald: Earned 50,000,000 credits (53,545,357)PSA Double Silver: Earned 200,000,000 credits (200,301,443)
Message 90984 - Posted: 7 Jan 2016 | 22:00:34 UTC

It seems that OCL4 Windows 32-bit executable has some problems which are throwing out all kinds of weird errors on XP: http://www.primegrid.com/result.php?resultid=677122985

Note the first line:

12:54:19 (3840): Can't open init data file - running in standalone mode

Of course init_data.xml is right here, but genefer cannot open/read it for unknown reason. I've tested it on THREE different XP systems and it always fails in same way.

I've tried to run it manually - no luck, always "cannot open file" error. When I've replaced OCL4 with standard genefer-CPU executable, it accepted same init_data.xml just fine.

I removed app_info.xml and switched back to default (OCL3) version, it's running fine.

The second problem is that it cannot rename checkpoint file (*.ckpt.new to *.ckpt). I think it does not mean anything by itself, it just another appearance of a global problem which caused first bug.

And the most weird thing that it seems to be running fine under Windows 7/64 bit. If you'll ask me "how could it be possible" - I don't know. May be it's something wrong with build options - e.g. different compilers/calling conventions were used to build different parts of project (e.g. boinc libraries vs. main application) - this will cause stack corruption. May be it's a global memory corruption.

Alas, I've realized this just a few hours before challenge. I have to undo all XP GPU systems to OCL3. Since OCL3 is almost 50% slower on 750TI, and consumes full CPU core, my challenge scores are expected to be lowered a lot...

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 91062 - Posted: 9 Jan 2016 | 22:33:30 UTC

I just noticed-my WR unit using OCL4 on a 980 (not ti) on a mac validated:

http://www.primegrid.com/workunit.php?wuid=455531628

74 hours, not blazing like that ti, but a heck of a lot faster than plain ocl.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91123 - Posted: 11 Jan 2016 | 21:59:09 UTC

WARNING!

If you are using app_info to run the new (build 978) version of OCL4 on GFN-16 (65536), the leading edge of GFN-16 is getting fairly close to the b-limit of that version of OCL4. The limit is 5,720,809 and the leading edge is currently at 5,580,264.

Any test you run above the b-limit will have 100% chance of failure.

I'm estimating there's only three or four days left before the leading edge surpasses the b-limit.

If you want to continue running GFN-16 you will need to switch to either the older (build 955) version of OCL4 or switch to OCL3. The server, of course, sends out OCL3.

The older 955 build of OCL4 is good up to b=9,260,772 and OCL3 is good up until b=11,863,283.
____________
My lucky number is 75898524288+1

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91132 - Posted: 12 Jan 2016 | 8:33:57 UTC - in response to Message 91123.

The older 955 build of OCL4 is good up to b=9,260,772 and OCL3 is good up until b=11,863,283.

Enjoying latest build so much, I do not poses build 955 for Windows anymore.
Where can I download it?

Any update on combined app progress?
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91138 - Posted: 12 Jan 2016 | 15:19:12 UTC - in response to Message 91123.

I'm estimating there's only three or four days left before the leading edge surpasses the b-limit.


I am revising this estimate upwards somewhat. Considering the smaller number of n=16 tasks being processed right now (they're not part of the challenge), it looks like we've got about 5 to 7 days left before hitting the "new" OCL4 limit. Sometime Sunday through Tuesday seems most likely.
____________
My lucky number is 75898524288+1

Iain BethuneProject donor
Honorary cruncher
Send message
Joined: 28 Jan 09
Posts: 1588
ID: 34775
Credit: 194,056,043
RAC: 0
321 LLR Gold: Earned 500,000 credits (597,175)Cullen LLR Amethyst: Earned 1,000,000 credits (1,355,381)ESP LLR Turquoise: Earned 5,000,000 credits (5,048,594)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,564,412)PPS LLR Amethyst: Earned 1,000,000 credits (1,025,115)PSP LLR Sapphire: Earned 20,000,000 credits (28,367,849)SoB LLR Jade: Earned 10,000,000 credits (16,215,329)SR5 LLR Turquoise: Earned 5,000,000 credits (5,189,992)SGS LLR Amethyst: Earned 1,000,000 credits (1,244,067)TRP LLR Turquoise: Earned 5,000,000 credits (5,169,405)Woodall LLR Amethyst: Earned 1,000,000 credits (1,070,956)321 Sieve Bronze: Earned 10,000 credits (20,003)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (200,371)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (11,645,025)PPS Sieve Turquoise: Earned 5,000,000 credits (7,536,532)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,017,144)TRP Sieve (suspended) Gold: Earned 500,000 credits (554,946)AP 26/27 Ruby: Earned 2,000,000 credits (3,577,848)GFN Emerald: Earned 50,000,000 credits (97,047,976)PSA Ruby: Earned 2,000,000 credits (4,606,694)
Message 91146 - Posted: 12 Jan 2016 | 18:32:04 UTC - in response to Message 91132.


Enjoying latest build so much, I do not poses build 955 for Windows anymore.
Where can I download it?

Any update on combined app progress?


GeneferOCL4 from revision 955 can be found here: https://www.assembla.com/spaces/genefer/subversion/source/955/trunk/bin/windows/geneferocl4_windows.exe (click on the 'Download' link on the right).

Re: the combined app, I have a working version containing OCL/OCL2/OCL3 and I will be posting binaries for testing after the challenge. OCL4 is still separate for now until development reaches the point where it makes sense to merge.

- Iain
____________
Twitter: IainBethune
Proud member of team "Aggie The Pew". Go Aggie!
3073428256125*2^1290000-1 is Prime!

recoil44Project donor
Avatar
Send message
Joined: 20 Dec 15
Posts: 167
ID: 433037
Credit: 411,347,492
RAC: 0
Discovered 1 mega primeFound 1 prime in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (5,019,377)Cullen LLR Turquoise: Earned 5,000,000 credits (5,064,170)ESP LLR Turquoise: Earned 5,000,000 credits (5,001,303)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,094,139)PPS LLR Turquoise: Earned 5,000,000 credits (6,021,807)PSP LLR Turquoise: Earned 5,000,000 credits (5,085,124)SoB LLR Turquoise: Earned 5,000,000 credits (5,042,051)SR5 LLR Turquoise: Earned 5,000,000 credits (5,077,248)SGS LLR Turquoise: Earned 5,000,000 credits (5,000,448)TRP LLR Turquoise: Earned 5,000,000 credits (5,003,404)Woodall LLR Turquoise: Earned 5,000,000 credits (5,035,352)321 Sieve Turquoise: Earned 5,000,000 credits (5,061,436)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,012,148)PPS Sieve Double Bronze: Earned 100,000,000 credits (108,556,313)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,089,865)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,297,782)AP 26/27 Emerald: Earned 50,000,000 credits (50,153,415)GFN Double Bronze: Earned 100,000,000 credits (106,065,222)PSA Emerald: Earned 50,000,000 credits (67,666,890)
Message 91157 - Posted: 12 Jan 2016 | 21:37:54 UTC - in response to Message 91062.

I just noticed-my WR unit using OCL4 on a 980 (not ti) on a mac validated:

http://www.primegrid.com/workunit.php?wuid=455531628

74 hours, not blazing like that ti, but a heck of a lot faster than plain ocl.


New to PrimeGrid and was looking at using OCL4 but this confused me because that's only a 1/2 hour faster than my 780 with the default OCL.

Only run one WR so far and it's not validated so it might not be a fair comparison.
Am I missing something?

http://www.primegrid.com/workunit.php?wuid=460781639

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 91159 - Posted: 12 Jan 2016 | 21:49:24 UTC - in response to Message 91157.


New to PrimeGrid and was looking at using OCL4 but this confused me because that's only a 1/2 hour faster than my 780 with the default OCL.

Only run one WR so far and it's not validated so it might not be a fair comparison.
Am I missing something?

http://www.primegrid.com/workunit.php?wuid=460781639


Running straight ocl my 980 was running over 100 hours. Is yours in a mac (and is it a ti)?

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91161 - Posted: 12 Jan 2016 | 22:55:44 UTC - in response to Message 91157.

New to PrimeGrid and was looking at using OCL4 but this confused me because that's only a 1/2 hour faster than my 780 with the default OCL.

OCL4 is faster than OCL on Maxwell (900 series) but runs at about the same speed on Kepler (600 & 700) and is slower on Fermi (500).

recoil44Project donor
Avatar
Send message
Joined: 20 Dec 15
Posts: 167
ID: 433037
Credit: 411,347,492
RAC: 0
Discovered 1 mega primeFound 1 prime in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (5,019,377)Cullen LLR Turquoise: Earned 5,000,000 credits (5,064,170)ESP LLR Turquoise: Earned 5,000,000 credits (5,001,303)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,094,139)PPS LLR Turquoise: Earned 5,000,000 credits (6,021,807)PSP LLR Turquoise: Earned 5,000,000 credits (5,085,124)SoB LLR Turquoise: Earned 5,000,000 credits (5,042,051)SR5 LLR Turquoise: Earned 5,000,000 credits (5,077,248)SGS LLR Turquoise: Earned 5,000,000 credits (5,000,448)TRP LLR Turquoise: Earned 5,000,000 credits (5,003,404)Woodall LLR Turquoise: Earned 5,000,000 credits (5,035,352)321 Sieve Turquoise: Earned 5,000,000 credits (5,061,436)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,012,148)PPS Sieve Double Bronze: Earned 100,000,000 credits (108,556,313)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,089,865)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,297,782)AP 26/27 Emerald: Earned 50,000,000 credits (50,153,415)GFN Double Bronze: Earned 100,000,000 credits (106,065,222)PSA Emerald: Earned 50,000,000 credits (67,666,890)
Message 91169 - Posted: 13 Jan 2016 | 2:33:53 UTC - in response to Message 91159.


Running straight ocl my 980 was running over 100 hours. Is yours in a mac (and is it a ti)?


Not in a Mac, not a Ti. It's specifically an ASUS GTX780-DC2OC-3GD5 running it's stock (factory OC) clocks.

recoil44Project donor
Avatar
Send message
Joined: 20 Dec 15
Posts: 167
ID: 433037
Credit: 411,347,492
RAC: 0
Discovered 1 mega primeFound 1 prime in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (5,019,377)Cullen LLR Turquoise: Earned 5,000,000 credits (5,064,170)ESP LLR Turquoise: Earned 5,000,000 credits (5,001,303)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,094,139)PPS LLR Turquoise: Earned 5,000,000 credits (6,021,807)PSP LLR Turquoise: Earned 5,000,000 credits (5,085,124)SoB LLR Turquoise: Earned 5,000,000 credits (5,042,051)SR5 LLR Turquoise: Earned 5,000,000 credits (5,077,248)SGS LLR Turquoise: Earned 5,000,000 credits (5,000,448)TRP LLR Turquoise: Earned 5,000,000 credits (5,003,404)Woodall LLR Turquoise: Earned 5,000,000 credits (5,035,352)321 Sieve Turquoise: Earned 5,000,000 credits (5,061,436)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,012,148)PPS Sieve Double Bronze: Earned 100,000,000 credits (108,556,313)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,089,865)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,297,782)AP 26/27 Emerald: Earned 50,000,000 credits (50,153,415)GFN Double Bronze: Earned 100,000,000 credits (106,065,222)PSA Emerald: Earned 50,000,000 credits (67,666,890)
Message 91172 - Posted: 13 Jan 2016 | 3:03:38 UTC - in response to Message 91161.

OCL4 is faster than OCL on Maxwell (900 series) but runs at about the same speed on Kepler (600 & 700) and is slower on Fermi (500).


OK, thanks. Probably won't mess with it then.

Also, did some research and found that 980s have a 1:32 FP64 rate performance while 780s are 1:24. So everything makes sense now.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91296 - Posted: 17 Jan 2016 | 14:41:08 UTC

And I'm back from vacation! Yay.

Anyways, I decided to extensively try OCL4 on my Gtx 970. And I'm having some small issues running tasks and using Media Player Classic. Whenever I open a video file, the driver immediately crashes. And before anyone asks, I did remove the OC. I have to keep trying to open the file multiple times until it eventually works after 5~15ish tries.

Not sure if we can do anything about it on our side of the coding, but that doesn't happen with the other OCL apps, so I just wanted to point that out.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91309 - Posted: 18 Jan 2016 | 14:28:58 UTC

A new (final?) version of GeneferOCL4 can be downloaded from the repository for beta testing.

OCL4 can test any GFN with b <= 400,000,000.
A 2-prime NTT is selected if b is 'small' otherwise a 3-prime NTT is used.

The bounds (2/3 primes) are:

m = 32768, b = 8090445 m = 65536, b = 5720809 m = 131072, b = 4045223 m = 262144, b = 2860404 m = 524288, b = 2022611 m = 1048576, b = 1430202 m = 2097152, b = 1011306 m = 4194304, b = 715101

On a GTX680, mark scores are:
OCL = 67 OCL4 = 65 (2 primes) OCL4 = 38 (3 primes) OCL3 = 47 OCL2 = 15

The 2-prime version is as fast as the previous version (slightly slower because error detection was improved).
OCL4 is faster than OCL3 for 'small' b, faster than OCL2 for any b.

In practice, OCL4 is the fastest GPU app for n=15, n=17 low, n=17 mega, n=18, n=19 and n=20.
For n=16, OCL3 or old-old-OCL4 are more efficient.
For n=21 and n=22, OCL or OCL4 selection depends on hardware.

The combined app will optimize n=16, 21 and 22 cases.

The app was compiled with gcc. I hope that it resolves issues with XP.

Yves

Scott BrownProject donor
Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar
Send message
Joined: 17 Oct 05
Posts: 2165
ID: 1178
Credit: 8,777,295,508
RAC: 144,324
Discovered the World's First base 116 Generalized Cullen prime!!!Discovered 26 mega primesEliminated 7 conjecture "k"sDiscovered 1 Sophie Germain pairDiscovered 2 Fermat divisors2012 Tour de Primes highest prime count2012 Tour de Primes most Mountain Stage primes2015 Tour de Primes highest prime count2016 Tour de Primes highest prime countFound 23 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 2 primes in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes highest prime countFound 22 primes in the 2019 Tour de Primes2020 Tour de Primes highest prime scoreFound 21 primes in the 2020 Tour de PrimesFound 4 mega primes in the 2020 Tour de Primes321 LLR Double Silver: Earned 200,000,000 credits (201,403,028)Cullen LLR Double Bronze: Earned 100,000,000 credits (188,339,206)ESP LLR Double Silver: Earned 200,000,000 credits (203,249,784)Generalized Cullen/Woodall LLR Double Bronze: Earned 100,000,000 credits (109,580,172)PPS LLR Double Gold: Earned 500,000,000 credits (647,115,820)PSP LLR Double Bronze: Earned 100,000,000 credits (126,982,721)SoB LLR Double Bronze: Earned 100,000,000 credits (135,747,083)SR5 LLR Double Silver: Earned 200,000,000 credits (214,194,272)SGS LLR Double Silver: Earned 200,000,000 credits (200,731,695)TPS LLR (retired) Silver: Earned 100,000 credits (235,439)TRP LLR Double Silver: Earned 200,000,000 credits (201,215,056)Woodall LLR Double Bronze: Earned 100,000,000 credits (101,447,725)321 Sieve Double Silver: Earned 200,000,000 credits (235,451,253)Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (83,794,448)Generalized Cullen/Woodall Sieve (suspended) Double Silver: Earned 200,000,000 credits (285,139,652)PPS Sieve Double Ruby: Earned 2,000,000,000 credits (2,728,039,937)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Double Silver: Earned 200,000,000 credits (203,523,358)TRP Sieve (suspended) Double Silver: Earned 200,000,000 credits (201,489,157)AP 26/27 Double Silver: Earned 200,000,000 credits (385,720,907)GFN Double Amethyst: Earned 1,000,000,000 credits (1,984,821,252)PSA Double Silver: Earned 200,000,000 credits (259,058,048)
Message 91310 - Posted: 18 Jan 2016 | 14:45:26 UTC
Last modified: 18 Jan 2016 | 15:02:04 UTC

On a GTX660-OEM, mark scores are:

OCL = 45 OCL4 = 44 (2 primes) OCL4 = 27 (3 primes) OCL3 = 31 OCL2 = 10



EDIT:
On a GTX660Ti, mark scores are:
OCL = 56 OCL4 = 50 (2 primes) OCL4 = 29 (3 primes) OCL3 = 38 OCL2 = 13

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91311 - Posted: 18 Jan 2016 | 14:47:36 UTC - in response to Message 90802.
Last modified: 18 Jan 2016 | 14:54:48 UTC

Wow, huge increase in b limit, beyond imagination.
And just in time for GFN16.

EDIT: Written too soon.

Testing 5645768^65536+1...
Using OCL4 transform

Starting initialization...
Initialization complete (0.121 seconds).
Estimated time remaining for 5645768^65536+1 is 0:02:28
Testing 5645768^65536+1... 1376256 steps to go (0:02:19 remaining)
maxErr exceeded for 5645768^65536+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations



Fury Nano

OCL: 131
OCL3: 70
OCL4 build 955: 86
OCL4 build 978: 107
OCL4 build 1004: 103

Running benchmarks for transform implementation "OCL4"

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driv
er '1912.5 (VM)'.

2199064^8192+1 Time: 82.9 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 72.2 us/mul. Err: 1.0000 102481 digits
1471094^32768+1 Time: 81.7 us/mul. Err: 1.0000 202102 digits
1203210^65536+1 Time: 90.2 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 140 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 221 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 505 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 1.01 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 2.01 ms/mul. Err: 1.0000 11836006 digits
360204^4194304+1 Time: 3.93 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 8.11 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 103.

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

eXaPower
Send message
Joined: 30 Sep 13
Posts: 122
ID: 259902
Credit: 1,461,890,363
RAC: 1,328
Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Amethyst: Earned 1,000,000 credits (1,031,618)Cullen LLR Gold: Earned 500,000 credits (620,013)ESP LLR Amethyst: Earned 1,000,000 credits (1,043,069)Generalized Cullen/Woodall LLR Gold: Earned 500,000 credits (585,120)PPS LLR Turquoise: Earned 5,000,000 credits (9,238,059)PSP LLR Gold: Earned 500,000 credits (506,876)SoB LLR Silver: Earned 100,000 credits (379,643)SR5 LLR Turquoise: Earned 5,000,000 credits (5,022,053)SGS LLR Amethyst: Earned 1,000,000 credits (1,006,021)TRP LLR Ruby: Earned 2,000,000 credits (2,403,013)Woodall LLR Gold: Earned 500,000 credits (509,918)321 Sieve Ruby: Earned 2,000,000 credits (2,035,186)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,122,369)PPS Sieve Double Gold: Earned 500,000,000 credits (758,913,230)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,058,245)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (3,185,685)AP 26/27 Turquoise: Earned 5,000,000 credits (9,901,307)GFN Double Gold: Earned 500,000,000 credits (656,149,179)
Message 91315 - Posted: 18 Jan 2016 | 15:09:27 UTC

A new (final?) version of GeneferOCL4 can be downloaded from the repository for beta testing.


OCL4 build 1004 benchmark 2 primes vs. 3 primes:

GT 630 @ 900MHz on PCIe 2.0 x1:


Generalized Fermat Number Bench
2199064^8192+1 Time: 233 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 267 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 390 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 742 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 1.73 ms/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 3.23 ms/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 6.33 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 12.2 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 24.4 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 47.9 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 117 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 8.

100000000^8192+1 Time: 259 us/mul. Err: 0.0000 65537 digits
100000000^16384+1 Time: 369 us/mul. Err: 0.0000 131073 digits
100000000^32768+1 Time: 567 us/mul. Err: 0.0000 262145 digits
100000000^65536+1 Time: 1.13 ms/mul. Err: 0.0000 524289 digits
100000000^131072+1 Time: 2.61 ms/mul. Err: 0.0000 1048577 digits
100000000^262144+1 Time: 4.9 ms/mul. Err: 0.0000 2097153 digits
100000000^524288+1 Time: 10 ms/mul. Err: 0.0000 4194305 digits
100000000^1048576+1 Time: 19.9 ms/mul. Err: 0.0000 8388609 digits
100000000^2097152+1 Time: 39.6 ms/mul. Err: 0.0000 16777217 digits
100000000^4194304+1 Time: 81.5 ms/mul. Err: 0.0000 33554433 digits
100000000^8388608+1 Time: 192 ms/mul. Err: 0.0000 67108865 digits
Genefer Mark = 5.

GTX 750 @ 1.4GHz on PCIe 3.0 x4:


Generalized Fermat Number Bench
2199064^8192+1 Time: 71 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 77.2 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 103 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 158 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 313 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 583 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 1.17 ms/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 2.33 ms/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 4.69 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 9.39 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 20.4 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 43.

100000000^8192+1 Time: 74.2 us/mul. Err: 0.0000 65537 digits
100000000^16384+1 Time: 92.9 us/mul. Err: 0.0000 131073 digits
100000000^32768+1 Time: 138 us/mul. Err: 0.0000 262145 digits
100000000^65536+1 Time: 263 us/mul. Err: 0.0000 524289 digits
100000000^131072+1 Time: 489 us/mul. Err: 0.0000 1048577 digits
100000000^262144+1 Time: 945 us/mul. Err: 0.0000 2097153 digits
100000000^524288+1 Time: 1.89 ms/mul. Err: 0.0000 4194305 digits
100000000^1048576+1 Time: 3.79 ms/mul. Err: 0.0000 8388609 digits
100000000^2097152+1 Time: 7.64 ms/mul. Err: 0.0000 16777217 digits
100000000^4194304+1 Time: 15.4 ms/mul. Err: 0.0000 33554433 digits
100000000^8388608+1 Time: 32.7 ms/mul. Err: 0.0000 67108865 digits
Genefer Mark = 26.

GTX 970 @ 1.5GHz on PCIe 3.0 x4:


Generalized Fermat Number Bench
2199064^8192+1 Time: 85.4 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 84.9 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 92.7 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 109 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 182 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 295 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 541 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 1e+003 us/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 1.9 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 3.68 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 8.8 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 104.

100000000^8192+1 Time: 87.9 us/mul. Err: 0.0000 65537 digits
100000000^16384+1 Time: 93.1 us/mul. Err: 0.0000 131073 digits
100000000^32768+1 Time: 101 us/mul. Err: 0.0000 262145 digits
100000000^65536+1 Time: 161 us/mul. Err: 0.0000 524289 digits
100000000^131072+1 Time: 252 us/mul. Err: 0.0000 1048577 digits
100000000^262144+1 Time: 452 us/mul. Err: 0.0000 2097153 digits
100000000^524288+1 Time: 833 us/mul. Err: 0.0000 4194305 digits
100000000^1048576+1 Time: 1.58 ms/mul. Err: 0.0000 8388609 digits
100000000^2097152+1 Time: 3.07 ms/mul. Err: 0.0000 16777217 digits
100000000^4194304+1 Time: 6.13 ms/mul. Err: 0.0000 33554433 digits
100000000^8388608+1 Time: 14.3 ms/mul. Err: 0.0000 67108865 digits
Genefer Mark = 64.

GTX 970 @ 1.5GHz on PCIe 3.0 x8:


Generalized Fermat Number Bench
2199064^8192+1 Time: 64.1 us/mul. Err: 0.0000 51956 digits
1798620^16384+1 Time: 61.9 us/mul. Err: 0.0000 102481 digits
1471094^32768+1 Time: 65.4 us/mul. Err: 0.0000 202102 digits
1203210^65536+1 Time: 95.3 us/mul. Err: 0.0000 398482 digits
984108^131072+1 Time: 155 us/mul. Err: 0.0000 785521 digits
804904^262144+1 Time: 267 us/mul. Err: 0.0000 1548156 digits
658332^524288+1 Time: 502 us/mul. Err: 0.0000 3050541 digits
538452^1048576+1 Time: 961 us/mul. Err: 0.0000 6009544 digits
440400^2097152+1 Time: 1.92 ms/mul. Err: 0.0000 11836006 digits
360204^4194304+1 Time: 3.67 ms/mul. Err: 0.0000 23305854 digits
294612^8388608+1 Time: 8.64 ms/mul. Err: 0.0000 45879398 digits
Genefer Mark = 105.

100000000^8192+1 Time: 65.3 us/mul. Err: 0.0000 65537 digits
100000000^16384+1 Time: 76.2 us/mul. Err: 0.0000 131073 digits
100000000^32768+1 Time: 85.3 us/mul. Err: 0.0000 262145 digits
100000000^65536+1 Time: 144 us/mul. Err: 0.0000 524289 digits
100000000^131072+1 Time: 225 us/mul. Err: 0.0000 1048577 digits
100000000^262144+1 Time: 420 us/mul. Err: 0.0000 2097153 digits
100000000^524288+1 Time: 797 us/mul. Err: 0.0000 4194305 digits
100000000^1048576+1 Time: 1.53 ms/mul. Err: 0.0000 8388609 digits
100000000^2097152+1 Time: 3.04 ms/mul. Err: 0.0000 16777217 digits
100000000^4194304+1 Time: 6.05 ms/mul. Err: 0.0000 33554433 digits
100000000^8388608+1 Time: 14 ms/mul. Err: 0.0000 67108865 digits
Genefer Mark = 65.

Scott BrownProject donor
Volunteer moderator
Project administrator
Volunteer tester
Project scientist
Avatar
Send message
Joined: 17 Oct 05
Posts: 2165
ID: 1178
Credit: 8,777,295,508
RAC: 144,324
Discovered the World's First base 116 Generalized Cullen prime!!!Discovered 26 mega primesEliminated 7 conjecture "k"sDiscovered 1 Sophie Germain pairDiscovered 2 Fermat divisors2012 Tour de Primes highest prime count2012 Tour de Primes most Mountain Stage primes2015 Tour de Primes highest prime count2016 Tour de Primes highest prime countFound 23 primes in the 2018 Tour de PrimesFound 1 mega prime in the 2018 Tour de PrimesFound 2 primes in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes highest prime countFound 22 primes in the 2019 Tour de Primes2020 Tour de Primes highest prime scoreFound 21 primes in the 2020 Tour de PrimesFound 4 mega primes in the 2020 Tour de Primes321 LLR Double Silver: Earned 200,000,000 credits (201,403,028)Cullen LLR Double Bronze: Earned 100,000,000 credits (188,339,206)ESP LLR Double Silver: Earned 200,000,000 credits (203,249,784)Generalized Cullen/Woodall LLR Double Bronze: Earned 100,000,000 credits (109,580,172)PPS LLR Double Gold: Earned 500,000,000 credits (647,115,820)PSP LLR Double Bronze: Earned 100,000,000 credits (126,982,721)SoB LLR Double Bronze: Earned 100,000,000 credits (135,747,083)SR5 LLR Double Silver: Earned 200,000,000 credits (214,194,272)SGS LLR Double Silver: Earned 200,000,000 credits (200,731,695)TPS LLR (retired) Silver: Earned 100,000 credits (235,439)TRP LLR Double Silver: Earned 200,000,000 credits (201,215,056)Woodall LLR Double Bronze: Earned 100,000,000 credits (101,447,725)321 Sieve Double Silver: Earned 200,000,000 credits (235,451,253)Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (83,794,448)Generalized Cullen/Woodall Sieve (suspended) Double Silver: Earned 200,000,000 credits (285,139,652)PPS Sieve Double Ruby: Earned 2,000,000,000 credits (2,728,039,937)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Double Silver: Earned 200,000,000 credits (203,523,358)TRP Sieve (suspended) Double Silver: Earned 200,000,000 credits (201,489,157)AP 26/27 Double Silver: Earned 200,000,000 credits (385,720,907)GFN Double Amethyst: Earned 1,000,000,000 credits (1,984,821,252)PSA Double Silver: Earned 200,000,000 credits (259,058,048)
Message 91316 - Posted: 18 Jan 2016 | 15:12:53 UTC - in response to Message 91311.
Last modified: 18 Jan 2016 | 15:13:27 UTC


EDIT: Written too soon.
Testing 5645768^65536+1...
Using OCL4 transform

Starting initialization...
Initialization complete (0.121 seconds).
Estimated time remaining for 5645768^65536+1 is 0:02:28
Testing 5645768^65536+1... 1376256 steps to go (0:02:19 remaining)
maxErr exceeded for 5645768^65536+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations



I don't get the same error on my GTX 660-OEM:
Testing 5645768^65536+1... Using OCL4 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 660', version 'OpenCL 1.2 CUDA' and driver '355.98'. The checkpoint doesn't match current test: 5645768^65536+1 != 1651902^16384+1. Current test will be restarted Starting initialization... Initialization complete (0.159 seconds). Estimated time remaining for 5645768^65536+1 is 0:04:00 5645768^65536+1 is a probable prime. (442481 digits) (err = 0.0000) (time = 0:03 :30) 10:11:08

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91317 - Posted: 18 Jan 2016 | 15:18:33 UTC - in response to Message 91316.

Testing 5645768^65536+1...
Using OCL4 transform

Starting initialization...
Initialization complete (0.121 seconds).
Estimated time remaining for 5645768^65536+1 is 0:02:28
Testing 5645768^65536+1... 1376256 steps to go (0:02:19 remaining)
maxErr exceeded for 5645768^65536+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations

I don't get the same error on my GTX 660-OEM:

A bug on ATI?

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 680', version 'OpenCL 1.2 CUDA' and driver '361.43'. Testing 5645768^65536+1... Using OCL4 transform Starting initialization... Initialization complete (0.179 seconds). Estimated time remaining for 5645768^65536+1 is 0:02:48 5645768^65536+1 is a probable prime. (442481 digits) (err = 0.0000) (time = 0:02:38) 16:15:53

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91327 - Posted: 18 Jan 2016 | 16:53:25 UTC - in response to Message 91317.
Last modified: 18 Jan 2016 | 17:03:10 UTC

Benchmarks for the GTX 970 at Factory OCs. Overall, slight slowdown from old_new_ocl4 (can we start calling these "ocl4 ver 1...3" plz?), though it's still much better than OCL3. Also, twice the scores for OCL2, at a much higher B limit. Congrats!

OCL = 70 OCL2 = 30 OCL3 = 82 OCL4 - V1 = 71 OCL4 - V2 = 103 OCL4 - V3 = 101 - 62


And we even get the bonus of next no CPU usage, meaning we can dedicate our cores to those SoB tasks! Time to get back to GFN Prime Crunching Master Race.......


A bug on ATI?

Maybe. I've experienced the same problem on my A6-3500 iGPU, "Max Err 1.000 > 0.500". My GTX 970, on the other hand, crunches it just fine, saying probable prime with Max Err 0.00

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91328 - Posted: 18 Jan 2016 | 17:01:02 UTC

On my GTX 580, I have run benchmarks for CUDA, OCL, OCL2, OCL3, OCL4 (v955 aka "old old"), OCL4 (v978 aka "old"), OCL4 low b (v 1004 aka "new"), OCL4 high b (also v1004/new), and the combined OCL app using the OCL, OCL2, and OCl3 transforms.

I'm not sure why, but the combined OCL app scored about 1% better than the discreet OCL and OCL3 apps, and about half a percent better than the discreet OCL2 app. It's possible that this is just experimental error.

Comparing the new OCL4 low-b transform against OCL, it was about 30% slower. Maxwell GPUs should see very different results here, of course.

Compared to OCL3, it's about 26% faster.

Compared to the old-old OCL4, it's about 10% faster.

It's about 4% slower than the 'old' OCL4 -- I suspect this is the effect of the extra error checking.

Comparing the new OCL4 high-b transform to OCL2, it's about 58% faster. That's a HUGE improvement. So much of an improvement that the new OCL4 high-b transform is only 17% slower than OCL3!

Over all, the high-b transform is about 34% slower than the low-b tranform.

(Numbers reflect n=32768 through n=4194304.)

____________
My lucky number is 75898524288+1

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91329 - Posted: 18 Jan 2016 | 17:07:30 UTC

Using just the Genefer Mark scores, on a GTX 580:

CUDA: 54
OCL: 72
OCL2: 20
OCL3: 39
OCL4 (955/old-old): 45
OCL4 (978/old): 53
OCL4 low-b: 51
OCL4 high-b: 33
Combined OCL: 73
Combined OCL2: 20
Combined OCL3: 40

____________
My lucky number is 75898524288+1

Profile GrebulonerProject donor
Volunteer tester
Avatar
Send message
Joined: 2 Nov 09
Posts: 359
ID: 49572
Credit: 1,734,244,065
RAC: 15,400
Discovered 1 mega primeFound 2 primes in the 2018 Tour de PrimesFound 4 primes in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,176,149)Cullen LLR Jade: Earned 10,000,000 credits (10,858,937)ESP LLR Jade: Earned 10,000,000 credits (11,401,438)Generalized Cullen/Woodall LLR Jade: Earned 10,000,000 credits (11,458,304)PPS LLR Sapphire: Earned 20,000,000 credits (33,262,524)PSP LLR Jade: Earned 10,000,000 credits (12,641,950)SoB LLR Sapphire: Earned 20,000,000 credits (20,393,852)SR5 LLR Sapphire: Earned 20,000,000 credits (21,058,593)SGS LLR Jade: Earned 10,000,000 credits (15,790,620)TRP LLR Sapphire: Earned 20,000,000 credits (20,191,201)Woodall LLR Jade: Earned 10,000,000 credits (13,066,992)321 Sieve Emerald: Earned 50,000,000 credits (55,630,889)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,178,073)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (56,046,594)PPS Sieve Double Gold: Earned 500,000,000 credits (521,014,891)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (9,468,384)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,076,645)AP 26/27 Double Gold: Earned 500,000,000 credits (500,106,006)GFN Double Silver: Earned 200,000,000 credits (251,361,687)PSA Double Bronze: Earned 100,000,000 credits (126,200,096)
Message 91333 - Posted: 18 Jan 2016 | 17:51:21 UTC

GTX980ti@1365core/6610mem:

OCL4v1: 122
OCL4v2: 160
OCL4v3low: 158
OCL4v3high: 98
____________
Eating more cheese on Thursdays.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91349 - Posted: 19 Jan 2016 | 10:06:51 UTC - in response to Message 91311.

Testing 5645768^65536+1...
Using OCL4 transform

Starting initialization...
Initialization complete (0.121 seconds).
Estimated time remaining for 5645768^65536+1 is 0:02:28
Testing 5645768^65536+1... 1376256 steps to go (0:02:19 remaining)
maxErr exceeded for 5645768^65536+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations

Please, could you check these primes on ATI:

    390931984 128
    357739396 256
    170093764 512
    226442304 1024
    332114176 2048


and run the integrated test.
Thanks, Yves

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91351 - Posted: 19 Jan 2016 | 11:36:15 UTC - in response to Message 91349.
Last modified: 19 Jan 2016 | 11:40:18 UTC

Please, could you check these primes on ATI:

    390931984 128
    357739396 256
    170093764 512
    226442304 1024
    332114176 2048


and run the integrated test.



Sure, run on AMD 280X, latest drivers, Win7x64
Batch file:
    geneferocl4_windows_v1004.exe -q "390931984^128+1"
    geneferocl4_windows_v1004.exe -q "357739396^256+1"
    geneferocl4_windows_v1004.exe -q "170093764^512+1"
    geneferocl4_windows_v1004.exe -q "226442304^1024+1"
    geneferocl4_windows_v1004.exe -q "332114176^2048+1"


Results:

    390931984^128+1 is a probable prime. (1100 digits) (err = 0.0000) (time = 0:00:03) 12:34:19
    357739396^256+1 is a probable prime. (2190 digits) (err = 0.0000) (time = 0:00:04) 12:34:23
    170093764^512+1 is a probable prime. (4215 digits) (err = 0.0000) (time = 0:00:04) 12:34:27
    226442304^1024+1 is a probable prime. (8556 digits) (err = 0.0000) (time = 0:00:06) 12:34:33
    332114176^2048+1 is a probable prime. (17452 digits) (err = 0.0000) (time = 0:00:09) 12:34:42



EDIT: Same results on Fury Nano.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

288larssonProject donor
Volunteer tester
Avatar
Send message
Joined: 17 Apr 10
Posts: 131
ID: 58815
Credit: 3,869,450,212
RAC: 54,698
Discovered 35 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 2 mega primes in the 2018 Tour de Primes2019 Tour de Primes highest prime scoreFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Emerald: Earned 50,000,000 credits (60,169,666)Cullen LLR Jade: Earned 10,000,000 credits (19,326,316)ESP LLR Emerald: Earned 50,000,000 credits (50,462,437)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (51,436,736)PPS LLR Double Gold: Earned 500,000,000 credits (604,509,707)PSP LLR Emerald: Earned 50,000,000 credits (55,832,054)SoB LLR Emerald: Earned 50,000,000 credits (60,176,718)SR5 LLR Sapphire: Earned 20,000,000 credits (36,959,361)SGS LLR Jade: Earned 10,000,000 credits (19,111,441)TRP LLR Emerald: Earned 50,000,000 credits (60,430,522)Woodall LLR Jade: Earned 10,000,000 credits (19,268,740)321 Sieve Turquoise: Earned 5,000,000 credits (5,015,934)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,081,963)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (13,723,448)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,127,295,825)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (14,372,149)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (11,393,528)AP 26/27 Sapphire: Earned 20,000,000 credits (32,865,322)GFN Double Amethyst: Earned 1,000,000,000 credits (1,394,526,669)PSA Double Silver: Earned 200,000,000 credits (226,139,676)
Message 91354 - Posted: 19 Jan 2016 | 12:12:08 UTC - in response to Message 91349.

Hi run the integrated test.

F:\Diverse\pippis\boinc\prp\genefer\Genefer>geneferocl4_1004.exe -t
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_1004.exe -t

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 100014^32+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'.

Starting initialization...
Initialization complete (0.001 seconds).
Testing 100014^32+1... 530 steps to go
maxErr exceeded for 100014^32+1, 1.0000 > 0.4500 during final check

F:\Diverse\pippis\boinc\prp\genefer\Genefer>geneferocl4_1004.exe -r
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_1004.exe -r

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 10234^64+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'.


The checkpoint doesn't match current test: 10234^64+1 != 100014^32+1. Current test will be restarted
Starting initialization...
Initialization complete (0.001 seconds).
10234^64+1 is composite. (RES=e99e17786d3bacc2) (257 digits) (err = 0.0000) (time = 0:00:03) 12:22:41

Testing 10032^128+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 10032^128+1 is 0:00:00
10032^128+1 is composite. (RES=1d9767ae8d652d65) (513 digits) (err = 0.0000) (time = 0:00:03) 12:22:44

Testing 584328^256+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 584328^256+1 is 0:00:00
584328^256+1 is composite. (RES=32798be013527165) (1477 digits) (err = 0.0000) (time = 0:00:04) 12:22:48

Testing 419000^512+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 419000^512+1 is 0:00:00
419000^512+1 is composite. (RES=fee1d2c0894a662f) (2879 digits) (err = 0.0000) (time = 0:00:03) 12:22:51

Testing 352220^1024+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 352220^1024+1 is 0:00:01
352220^1024+1 is composite. (RES=6c7ebc76febe0583) (5680 digits) (err = 0.0000) (time = 0:00:04) 12:22:56

Testing 366672^2048+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 366672^2048+1 is 0:00:02

maxErr exceeded for 366672^2048+1, 1.0000 > 0.4500 during final check

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91355 - Posted: 19 Jan 2016 | 13:12:50 UTC - in response to Message 91354.

Hi run the integrated test.

F:\Diverse\pippis\boinc\prp\genefer\Genefer>geneferocl4_1004.exe -t
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_1004.exe -t

Priority change succeeded.
Running tests for transform implementation "OCL4"

...

Testing 366672^2048+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 366672^2048+1 is 0:00:02

maxErr exceeded for 366672^2048+1, 1.0000 > 0.4500 during final check


Is that repeatable?
____________
My lucky number is 75898524288+1

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91356 - Posted: 19 Jan 2016 | 13:24:19 UTC - in response to Message 91355.
Last modified: 19 Jan 2016 | 13:26:44 UTC

Is that repeatable?


Yes, it is on my 280X as well.

Command line: geneferocl4_windows_v1004.exe -t

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 100014^32+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Tahiti', version 'OpenCL 1.2 AMD-APP (1912.5)' and dr
iver '1912.5 (VM)'.

Starting initialization...
Initialization complete (0.003 seconds).
Testing 100014^32+1... 530 steps to go
maxErr exceeded for 100014^32+1, 1.0000 > 0.4500 during final check




c:\_PG>geneferocl4_windows_v1004.exe -r
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows_v1004.exe -r

...

Testing 366672^2048+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.007 seconds).
Estimated time remaining for 366672^2048+1 is 0:00:02

maxErr exceeded for 366672^2048+1, 1.0000 > 0.4500 during final check
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91362 - Posted: 19 Jan 2016 | 14:56:24 UTC - in response to Message 91356.

geneferocl4_windows_noErr.exe can be downloaded from the repository for test on AMD cards.
I removed the error detection because I think that the bug may be in this part of the code.

Thanks.

PS: why so many Mersenne primes and so few Generalized Fermat? (https://primes.utm.edu/notes/faq/NextMersenne.html)

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91363 - Posted: 19 Jan 2016 | 15:03:15 UTC - in response to Message 91123.

WARNING!

If you are using app_info to run the new (build 978) version of OCL4 on GFN-16 (65536), the leading edge of GFN-16 is getting fairly close to the b-limit of that version of OCL4. The limit is 5,720,809 and the leading edge is currently at 5,580,264.

Any test you run above the b-limit will have 100% chance of failure.

I'm estimating there's only three or four days left before the leading edge surpasses the b-limit.

If you want to continue running GFN-16 you will need to switch to either the older (build 955) version of OCL4 or switch to OCL3. The server, of course, sends out OCL3.

The older 955 build of OCL4 is good up to b=9,260,772 and OCL3 is good up until b=11,863,283.


The leading edge just surpassed OCL4 v978's b-limit. If you're using anonymous platform to run OCL4 and are using build 978, it's going to stop working.

____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91366 - Posted: 19 Jan 2016 | 15:33:38 UTC - in response to Message 91355.
Last modified: 19 Jan 2016 | 15:34:28 UTC

Is that repeatable?

Yes it is. I've got the exact same results as Honza on my A6-3500 iGPU.

geneferocl4_windows_noErr.exe can be downloaded from the repository for test on AMD cards.
I removed the error detection because I think that the bug may be in this part of the code.

It seems to be running fine now.

Command line: geneferocl4_windows_noErr.exe -t
Priority change succeeded. Running tests for transform implementation "OCL4" Testing 100014^32+1... Using OCL4 transform Running on platform 'AMD Accelerated Parallel Processing', device 'BeaverCreek', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. The checkpoint doesn't match current test: 100014^32+1 != 366672^2048+1. Current test will be restarted Starting initialization... Initialization complete (0.002 seconds). 100014^32+1 is a probable prime. (161 digits) (err = 0.0000) (time = 0:00:06) 13:22:59 Testing 100234^64+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 100234^64+1 is 0:00:00 100234^64+1 is a probable prime. (321 digits) (err = 0.0000) (time = 0:00:06) 13:23:05 Testing 100032^128+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 100032^128+1 is 0:00:00 100032^128+1 is a probable prime. (641 digits) (err = 0.0000) (time = 0:00:05) 13:23:10 Testing 5684328^256+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 5684328^256+1 is 0:00:00 5684328^256+1 is a probable prime. (1730 digits) (err = 0.0000) (time = 0:00:06) 13:23:16 Testing 4619000^512+1... Using OCL4 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 4619000^512+1 is 0:00:00 4619000^512+1 is a probable prime. (3413 digits) (err = 0.0000) (time = 0:00:07) 13:23:23 Testing 3752220^1024+1... Using OCL4 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 3752220^1024+1 is 0:00:01 3752220^1024+1 is a probable prime. (6733 digits) (err = 0.0000) (time = 0:00:07) 13:23:30 Testing 3066672^2048+1... Using OCL4 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 3066672^2048+1 is 0:00:03 3066672^2048+1 is a probable prime. (13285 digits) (err = 0.0000) (time = 0:00:10) 13:23:40 Testing 2485064^4096+1... Using OCL4 transform Starting initialization... Initialization complete (0.005 seconds). Estimated time remaining for 2485064^4096+1 is 0:00:09 2485064^4096+1 is a probable prime. (26196 digits) (err = 0.0000) (time = 0:00:15) 13:23:55 Testing 2030234^8192+1... Using OCL4 transform Starting initialization... Initialization complete (0.015 seconds). Estimated time remaining for 2030234^8192+1 is 0:00:26 2030234^8192+1 is a probable prime. (51672 digits) (err = 0.0000) (time = 0:00:32) 13:24:27 Testing 1651902^16384+1... Using OCL4 transform Starting initialization... Initialization complete (0.037 seconds). Estimated time remaining for 1651902^16384+1 is 0:01:30 1651902^16384+1 is a probable prime. (101876 digits) (err = 0.0000) (time = 0:01:36) 13:26:03



Command line: geneferocl4_windows_noErr.exe -r
Priority change succeeded. Running tests for transform implementation "OCL4" Testing 10234^64+1... Using OCL4 transform Running on platform 'AMD Accelerated Parallel Processing', device 'BeaverCreek', version 'OpenCL 1.2 AMD-APP (1800.11)' and driver '1800.11 (VM)'. The checkpoint doesn't match current test: 10234^64+1 != 1277444^32768+1. Current test will be restarted Starting initialization... Initialization complete (0.002 seconds). 10234^64+1 is composite. (RES=e99e17786d3bacc2) (257 digits) (err = 0.0000) (time = 0:00:06) 13:27:25 Testing 10032^128+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 10032^128+1 is 0:00:00 10032^128+1 is composite. (RES=1d9767ae8d652d65) (513 digits) (err = 0.0000) (time = 0:00:06) 13:27:31 Testing 584328^256+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 584328^256+1 is 0:00:00 584328^256+1 is composite. (RES=32798be013527165) (1477 digits) (err = 0.0000) (time = 0:00:07) 13:27:38 Testing 419000^512+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 419000^512+1 is 0:00:00 419000^512+1 is composite. (RES=fee1d2c0894a662f) (2879 digits) (err = 0.0000) (time = 0:00:07) 13:27:45 Testing 352220^1024+1... Using OCL4 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 352220^1024+1 is 0:00:01 352220^1024+1 is composite. (RES=6c7ebc76febe0583) (5680 digits) (err = 0.0000) (time = 0:00:08) 13:27:53 Testing 366672^2048+1... Using OCL4 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 366672^2048+1 is 0:00:04 366672^2048+1 is composite. (RES=fd37a864102cf81e) (11396 digits) (err = 0.0000) (time = 0:00:09) 13:28:02 Testing 285064^4096+1... Using OCL4 transform Starting initialization... Initialization complete (0.005 seconds). Estimated time remaining for 285064^4096+1 is 0:00:07 285064^4096+1 is composite. (RES=f81ed37ab6df33f9) (22344 digits) (err = 0.0000) (time = 0:00:14) 13:28:16 Testing 230234^8192+1... Using OCL4 transform Starting initialization... Initialization complete (0.011 seconds). Estimated time remaining for 230234^8192+1 is 0:00:22 230234^8192+1 is composite. (RES=5972a302c255e081) (43927 digits) (err = 0.0000) (time = 0:00:30) 13:28:46 Testing 151902^16384+1... Using OCL4 transform Starting initialization... Initialization complete (0.032 seconds). Estimated time remaining for 151902^16384+1 is 0:01:17 151902^16384+1 is composite. (RES=0a1a98e69b17f1e9) (84895 digits) (err = 0.0000) (time = 0:01:25) 13:30:11

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91368 - Posted: 19 Jan 2016 | 17:14:06 UTC - in response to Message 91366.

It seems to be running fine now.

A new beta geneferocl4_windows.exe is available. I hope that AMD/ATI bug is fixed.

Yves

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91371 - Posted: 19 Jan 2016 | 18:23:12 UTC - in response to Message 91368.

A new beta geneferocl4_windows.exe is available. I hope that AMD/ATI bug is fixed.


c:\temp>geneferocl4_windows_v1006.exe -t
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows_v1006.exe -t

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 100014^32+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driv
er '1912.5 (VM)'.

Starting initialization...
Initialization complete (0.001 seconds).
Testing 100014^32+1... 530 steps to go
maxErr exceeded for 100014^32+1, 1.0000 > 0.4500 during final check

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91372 - Posted: 19 Jan 2016 | 18:25:07 UTC
Last modified: 19 Jan 2016 | 18:42:01 UTC

run on Fury Nano, latest drivers, Win7x64
geneferocl4_windows_v1006.exe -q "390931984^128+1"
geneferocl4_windows_v1006.exe -q "357739396^256+1"
geneferocl4_windows_v1006.exe -q "170093764^512+1"
geneferocl4_windows_v1006.exe -q "226442304^1024+1"
geneferocl4_windows_v1006.exe -q "332114176^2048+1"

Those tests are OK.

EDIT: Leading edge for GFN16 passed, but latest known prime for GFN16 did not.

Command line: geneferocl4_windows_v1006.exe -q 5738572^65536+1

Priority change succeeded.

Testing 5738572^65536+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driv
er '1912.5 (VM)'.

Starting initialization...
Initialization complete (0.135 seconds).
Estimated time remaining for 5738572^65536+1 is 0:04:12
5738572^65536+1 is composite. (RES=145e1cd1eb5c4c4e) (442945 digits) (err = 0.0000) (time = 0:04:16) 19:30:24

Command line: geneferocl4_windows_v1006.exe -q 5645768^65536+1

Priority change succeeded.

Testing 5645768^65536+1...
Using OCL4 transform

Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driv
er '1912.5 (VM)'.

Starting initialization...
Initialization complete (0.123 seconds).
Estimated time remaining for 5645768^65536+1 is 0:02:24
Testing 5645768^65536+1... 1376256 steps to go (0:02:13 remaining)
maxErr exceeded for 5645768^65536+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations

____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91375 - Posted: 19 Jan 2016 | 19:08:49 UTC - in response to Message 91372.

EDIT: Leading edge for GFN16 passed, but latest known prime for GFN16 did not.

A beta 3 (1007) geneferocl4_windows.exe is available.
Logically AMD/ATI bug is fixed... in practice, I'm still hoping (blind debug).

Yves

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91377 - Posted: 19 Jan 2016 | 19:20:18 UTC - in response to Message 91375.
Last modified: 19 Jan 2016 | 19:57:40 UTC

A beta 3 (1007) geneferocl4_windows.exe is available.
Logically AMD/ATI bug is fixed... in practice, I'm still hoping (blind debug).


-t test got maxerr on this test (smaller ones passed)

Testing 1277444^32768+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.038 seconds).
Estimated time remaining for 1277444^32768+1 is 0:00:51
Testing 1277444^32768+1... 262144 steps to go (0:00:21 remaining)
maxErr exceeded for 1277444^32768+1, 1.0000 > 0.4500


EDIT: latest known GFN16 prime test passed.
-r test up to GFN18 is OK.
70000^262144+1 is composite. (RES=fa15b4b858fd2ff0) (1270114 digits) (err = 0.0000) (time = 0:15:53) 20:45:10
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91378 - Posted: 19 Jan 2016 | 20:30:00 UTC - in response to Message 91377.

Testing 1277444^32768+1... 262144 steps to go (0:00:21 remaining)
maxErr exceeded for 1277444^32768+1, 1.0000 > 0.4500

EDIT: latest known GFN16 prime test passed.
-r test up to GFN18 is OK.
70000^262144+1 is composite. (RES=fa15b4b858fd2ff0) (1270114 digits) (err = 0.0000) (time = 0:15:53) 20:45:10

Only one test failed? Is that repeatable?

I'm testing HD Graphics

Running on platform 'Intel(R) OpenCL', device 'Intel(R) HD Graphics 4000', version 'OpenCL 1.2 ' and driver '10.18.10.4276'. Testing 5645768^65536+1... Using OCL4 transform Starting initialization... Initialization complete (0.259 seconds). Estimated time remaining for 5645768^65536+1 is 0:38:33 5645768^65536+1 is a probable prime. (442481 digits) (err = 0.0000) (time = 0:37:32) 20:51:40

Testing 1277444^32768+1... Using OCL4 transform Starting initialization... Initialization complete (0.079 seconds). Estimated time remaining for 1277444^32768+1 is 0:09:08 1277444^32768+1 is a probable prime. (200093 digits) (err = 0.0000) (time = 0:09:05) 21:27:54

No error until now.

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91381 - Posted: 19 Jan 2016 | 21:21:07 UTC - in response to Message 91378.
Last modified: 19 Jan 2016 | 21:21:23 UTC

Only one test failed? Is that repeatable?

Run on 280X passed.
Tested again on Fury Nano, this time passed.
Sorry, probably my bad.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

288larssonProject donor
Volunteer tester
Avatar
Send message
Joined: 17 Apr 10
Posts: 131
ID: 58815
Credit: 3,869,450,212
RAC: 54,698
Discovered 35 mega primesEliminated 1 conjecture "k"Found 3 primes in the 2018 Tour de PrimesFound 2 mega primes in the 2018 Tour de Primes2019 Tour de Primes highest prime scoreFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 3 primes in the 2020 Tour de PrimesFound 1 mega prime in the 2020 Tour de Primes321 LLR Emerald: Earned 50,000,000 credits (60,169,666)Cullen LLR Jade: Earned 10,000,000 credits (19,326,316)ESP LLR Emerald: Earned 50,000,000 credits (50,462,437)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (51,436,736)PPS LLR Double Gold: Earned 500,000,000 credits (604,509,707)PSP LLR Emerald: Earned 50,000,000 credits (55,832,054)SoB LLR Emerald: Earned 50,000,000 credits (60,176,718)SR5 LLR Sapphire: Earned 20,000,000 credits (36,959,361)SGS LLR Jade: Earned 10,000,000 credits (19,111,441)TRP LLR Emerald: Earned 50,000,000 credits (60,430,522)Woodall LLR Jade: Earned 10,000,000 credits (19,268,740)321 Sieve Turquoise: Earned 5,000,000 credits (5,015,934)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,081,963)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (13,723,448)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,127,295,825)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (14,372,149)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (11,393,528)AP 26/27 Sapphire: Earned 20,000,000 credits (32,865,322)GFN Double Amethyst: Earned 1,000,000,000 credits (1,394,526,669)PSA Double Silver: Earned 200,000,000 credits (226,139,676)
Message 91382 - Posted: 19 Jan 2016 | 21:26:47 UTC - in response to Message 91378.

I got no error with the -t test

Command line: geneferocl4_1007 -t Running on platform 'AMD Accelerated Parallel Processing', device 'Fiji', version 'OpenCL 2.0 AMD-APP (1912.5)' and driver '1912.5 (VM)'. Testing 475856^524288+1... Using OCL4 transform Starting initialization... Initialization complete (2.330 seconds). Estimated time remaining for 475856^524288+1 is 1:15:39 475856^524288+1 is a probable prime. (2976633 digits) (err = 0.0000) (time = 1:14:59) 22:04:32

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91402 - Posted: 20 Jan 2016 | 15:23:16 UTC

Tested 1007 with the A6-3500 iGPU. -t and -r all ran fine. The candidates that showed any problem before, all fine. Tested a few recent primes from n=15 and n=16, all green. I guess the bug is fixed now....

Also did some tests on the Intel HD 530 (Skylake's iGPU). While everything ran fine, it is sooooo slow..... like, 14min for an n=16 task. Even considering the new core optimization wasting almost 0 CPU time (so in theory it would be a plus to run iGPU tasks, even with them being so slow), I suppose the memory bottleneck for the actual CPU tasks makes it a big no no.

Still, the tests all tests validated up to n=16, so I guess the app is at least okay on this one.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91408 - Posted: 20 Jan 2016 | 17:34:27 UTC - in response to Message 91402.

If the last build runs fine, this is the release candidate. No further improvement is expected.

Profile DaveProject donor
Avatar
Send message
Joined: 13 Feb 12
Posts: 2829
ID: 130544
Credit: 954,747,840
RAC: 5,071
Found 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Turquoise: Earned 5,000,000 credits (6,042,688)Cullen LLR Turquoise: Earned 5,000,000 credits (5,617,541)ESP LLR Turquoise: Earned 5,000,000 credits (5,303,526)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,202,873)PPS LLR Turquoise: Earned 5,000,000 credits (6,100,156)PSP LLR Turquoise: Earned 5,000,000 credits (5,513,728)SoB LLR Turquoise: Earned 5,000,000 credits (8,851,996)SR5 LLR Turquoise: Earned 5,000,000 credits (5,199,189)SGS LLR Turquoise: Earned 5,000,000 credits (5,100,018)TRP LLR Turquoise: Earned 5,000,000 credits (5,405,027)Woodall LLR Turquoise: Earned 5,000,000 credits (5,491,055)321 Sieve Jade: Earned 10,000,000 credits (10,003,334)Cullen/Woodall Sieve (suspended) Silver: Earned 100,000 credits (268,250)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,502)PPS Sieve Double Silver: Earned 200,000,000 credits (306,761,000)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,133)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,000,970)AP 26/27 Double Silver: Earned 200,000,000 credits (200,128,500)GFN Double Bronze: Earned 100,000,000 credits (138,157,354)PSA Double Silver: Earned 200,000,000 credits (200,000,001)
Message 91411 - Posted: 20 Jan 2016 | 18:54:33 UTC

Woohoo!

Profile Rick Reynolds
Avatar
Send message
Joined: 13 Aug 10
Posts: 7880
ID: 65483
Credit: 5,312,080,766
RAC: 66,236
Discovered 2 mega primesFound 2 primes in the 2018 Tour de PrimesFound 2 primes in the 2019 Tour de PrimesFound 12 primes in the 2020 Tour de Primes321 LLR Jade: Earned 10,000,000 credits (10,244,909)Cullen LLR Jade: Earned 10,000,000 credits (10,194,451)ESP LLR Turquoise: Earned 5,000,000 credits (5,174,327)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (5,028,260)PPS LLR Emerald: Earned 50,000,000 credits (61,072,468)PSP LLR Turquoise: Earned 5,000,000 credits (5,151,448)SoB LLR Turquoise: Earned 5,000,000 credits (5,014,268)SR5 LLR Jade: Earned 10,000,000 credits (14,871,275)SGS LLR Turquoise: Earned 5,000,000 credits (7,860,522)TRP LLR Jade: Earned 10,000,000 credits (11,130,742)Woodall LLR Turquoise: Earned 5,000,000 credits (5,419,909)321 Sieve Double Bronze: Earned 100,000,000 credits (101,538,236)Cullen/Woodall Sieve (suspended) Double Bronze: Earned 100,000,000 credits (127,922,665)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (30,984,158)PPS Sieve Double Ruby: Earned 2,000,000,000 credits (2,840,173,136)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Turquoise: Earned 5,000,000 credits (8,134,964)TRP Sieve (suspended) Turquoise: Earned 5,000,000 credits (7,013,696)AP 26/27 Double Silver: Earned 200,000,000 credits (276,723,135)GFN Double Silver: Earned 200,000,000 credits (323,467,712)PSA Double Amethyst: Earned 1,000,000,000 credits (1,370,700,486)
Message 91412 - Posted: 20 Jan 2016 | 19:37:44 UTC - in response to Message 91408.

If the last build runs fine, this is the release candidate. No further improvement is expected.


And hopefully in production by Feb 1st :)

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91413 - Posted: 20 Jan 2016 | 20:35:33 UTC - in response to Message 91412.

If the last build runs fine, this is the release candidate. No further improvement is expected.


And hopefully in production by Feb 1st :)


I suspect that is incredibly unlikely. March 1st, maybe, but I wouldn't bet real money on it.

OCL4 won't be going into production as is. The next release will be the combined OCL app, which will incorporate OCL4 as one of the transforms.

OCL4 hasn't been incorporated into the combined app yet, since it's just today that it got to "release candidate" status. Once there's a combined app including OCL4, then we start testing it.

The combined app is brand new, and it will likely take several weeks to test adequately, if not several months. In part it depends on how many things need to be changed due to what we discover in testing, but also remember that there's a lot of builds of Genefer, and they all need to be tested on numerous hardware platforms. That takes time, and it also takes volunteers, since the admins and developers don't own every hardware platform we want to support.

My best guess is about 2 to 3 months.
____________
My lucky number is 75898524288+1

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91468 - Posted: 22 Jan 2016 | 19:02:40 UTC

Not sure if this is a problem with OCL4 or with my GPU. (I have been experiencing some intermittent Maxerr-type errors lately and am not sure what the cause is.)

I wanted to see what it would take to test a billion-digit number. Yes, there's a substantial prize for finding the first billion-digit prize. No, we're nowhere close to being able to search for one yet.

Since OCL4 can test b values up to 400,000,000, my first question was "How high does N have to be to represent a billion digit number?"

28221534^134217728+1 is 1,000,000,003 digits. Alternatively, 28221534^2^27+1 is probably a better way to write this number.

28221534^2^27+1 dies if you try to run it with OCL4. Memory allocation error.

Same thing with 28221534^2^26+1 and 28221534^2^25+1.

28221534^2^24+1 doesn't crash. It uses over 1 GB of memory on the host, and over 1 GB of memory on the GPU, and takes 18 minutes to perform the initialization -- meaning it took 18 minutes to repeatedly square 28221534 27 times to produce the 125 million digit number to be tested. Okay, so it finally started to run, using over a gigabyte of memory on both the CPU and GPU, and promptly estimated it would take 273 days to finish. So far so good. Then not so good:

C:\Temp\GFN\OCL4>geneferocl4_windows-1004.exe -q "28224534^16777216+1
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows-1004.exe -q 28224534^16777216+1

Priority change succeeded.

Testing 28224534^16777216+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'.

Starting initialization...
Initialization complete (1069.481 seconds).
Estimated time remaining for 28224534^16777216+1 is 6570:01:06
Testing 28224534^16777216+1... 415236096 steps to go (6575:42:32 remaining)

maxErr exceeded for 28224534^16777216+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations


I'm not sure if that was due to a problem on the GPU or a problem with OCL4. I'm not really interested in trying it again.
____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91470 - Posted: 22 Jan 2016 | 20:38:25 UTC - in response to Message 91468.
Last modified: 22 Jan 2016 | 20:40:09 UTC

28221534^2^24+1 doesn't crash. It uses over 1 GB of memory on the host, and over 1 GB of memory on the GPU, and takes 18 minutes to perform the initialization -- meaning it took 18 minutes to repeatedly square 28221534 27 times to produce the 125 million digit number to be tested. Okay, so it finally started to run, using over a gigabyte of memory on both the CPU and GPU, and promptly estimated it would take 273 days to finish. So far so good. Then not so good


HA! MINE ERRORS FASTER THAN YOU (is that supposed to be a good thing?)

Command line: geneferocl4_1007_windows -q 28224534^16777216+1 Priority change succeeded. Testing 28224534^16777216+1... Using OCL4 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'. Starting initialization... Initialization complete (726.859 seconds). Estimated time remaining for 28224534^16777216+1 is 3716:08:57 Testing 28224534^16777216+1... 415236096 steps to go (3723:38:49 remaining) maxErr exceeded for 28224534^16777216+1, 1.0000 > 0.4500 maxErr exceeded by all available transform implementations


Jokes aside, I decided to do a quick pass on n=23, as that's an actual possibility (at least it's more plausible than n=24 testing). Besides, I was asking myself "can a low b n=23 be faster than a high b n=22 due to using the slower 3 primes transform?"

Yves said the limit for n=22 is 715101. I upped to 715716 (kind of a funny number), which squared is 846. And surprise surprise, n=22 takes 147 hours and n=23 takes...... 215. Okay, maybe not a surprise. Yeah yeah, I get it, "always crunch the lowes n possible". Still, only being 1.5x the length is a nice deviation from my usual 3.5x rule of thumb.

Oh, and btw, no Max Err were found, at least not within the short spam of the test.

Van ZimmermanProject donor
Volunteer tester
Send message
Joined: 30 Aug 12
Posts: 1950
ID: 168418
Credit: 6,315,554,899
RAC: 0
Discovered the World's First GFN-20 prime!!!Discovered 2 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de Primes321 LLR Sapphire: Earned 20,000,000 credits (20,089,185)Cullen LLR Sapphire: Earned 20,000,000 credits (21,197,588)ESP LLR Sapphire: Earned 20,000,000 credits (20,089,407)Generalized Cullen/Woodall LLR Sapphire: Earned 20,000,000 credits (21,873,609)PPS LLR Sapphire: Earned 20,000,000 credits (21,918,476)PSP LLR Sapphire: Earned 20,000,000 credits (20,117,197)SoB LLR Sapphire: Earned 20,000,000 credits (26,779,370)SR5 LLR Sapphire: Earned 20,000,000 credits (20,278,667)SGS LLR Sapphire: Earned 20,000,000 credits (20,055,153)TRP LLR Sapphire: Earned 20,000,000 credits (20,619,871)Woodall LLR Sapphire: Earned 20,000,000 credits (20,394,541)321 Sieve Sapphire: Earned 20,000,000 credits (28,148,632)Generalized Cullen/Woodall Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,373,251)PPS Sieve Double Amethyst: Earned 1,000,000,000 credits (1,051,222,753)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Jade: Earned 10,000,000 credits (10,189,695)TRP Sieve (suspended) Jade: Earned 10,000,000 credits (10,102,079)AP 26/27 Emerald: Earned 50,000,000 credits (60,248,786)GFN Double Ruby: Earned 2,000,000,000 credits (4,703,540,875)PSA Double Bronze: Earned 100,000,000 credits (198,313,575)
Message 91476 - Posted: 22 Jan 2016 | 22:25:16 UTC

Errors out on a mac too.

./geneferocl4_macintel -q 28224534^16777216+1
geneferocl4 3.2.10-dev (Apple-x86/OpenCL-NTT-32-2/64-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: ./geneferocl4_macintel -q 28224534^16777216+1

Priority change succeeded.

Testing 28224534^16777216+1...
Using OCL4 transform

Running on platform 'Apple', device 'GeForce GTX 980', version 'OpenCL 1.2 ' and driver '10.5.2 346.02.03f01'.

Starting initialization...
Initialization complete (848.154 seconds).
Testing 28224534^16777216+1... 415243584 steps to go
maxErr exceeded for 28224534^16777216+1, 1.0000 > 0.4500
maxErr exceeded by all available transform implementations

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91546 - Posted: 25 Jan 2016 | 20:54:58 UTC

1) I see intermittent MaxErr errors using OCL. They don't happen very often, but they do occur frequently enough to be annoying on long tasks.

2) Anecdotally I've seen other people having similar problems.

3) It seems to me this problem is more likely to occur on larger Ns.

4) My silly test of n=24 seems to fail for everyone who has tested it.

I'm wondering if there's actually a bug in OCL -- maybe just OCL4 but perhaps all of the OCL apps -- that is affecting a lot of people, but not often enough that people are complaining about it.

If that's correct, then the n=24 test case, which seems to fail consistently (or at least more frequently), could perhaps be used to diagnose the problem.
____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91548 - Posted: 25 Jan 2016 | 21:42:51 UTC - in response to Message 91546.

1) I see intermittent MaxErr errors using OCL. They don't happen very often, but they do occur frequently enough to be annoying on long tasks.

2) Anecdotally I've seen other people having similar problems.

3) It seems to me this problem is more likely to occur on larger Ns.

4) My silly test of n=24 seems to fail for everyone who has tested it.

I'm wondering if there's actually a bug in OCL -- maybe just OCL4 but perhaps all of the OCL apps -- that is affecting a lot of people, but not often enough that people are complaining about it.

If that's correct, then the n=24 test case, which seems to fail consistently (or at least more frequently), could perhaps be used to diagnose the problem.

Quick question: if we get the same number written in for 2 different n, is the residue going to be the same? If so, we can test a ludicrous n=23 task against a known number to see if properly finishes the calculation. It seems that n= 24 very quickly reaches a point of max err; but what if n=23 has the same problem, it just takes longer to show up?

Also, has anyone tested WR tasks on the new app? I did an n=21 task and it validated, but besides that, didn't really mess with the higher ranges.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91551 - Posted: 25 Jan 2016 | 21:59:41 UTC - in response to Message 91548.

1) I see intermittent MaxErr errors using OCL. They don't happen very often, but they do occur frequently enough to be annoying on long tasks.

2) Anecdotally I've seen other people having similar problems.

3) It seems to me this problem is more likely to occur on larger Ns.

4) My silly test of n=24 seems to fail for everyone who has tested it.

I'm wondering if there's actually a bug in OCL -- maybe just OCL4 but perhaps all of the OCL apps -- that is affecting a lot of people, but not often enough that people are complaining about it.

If that's correct, then the n=24 test case, which seems to fail consistently (or at least more frequently), could perhaps be used to diagnose the problem.

Quick question: if we get the same number written in for 2 different n, is the residue going to be the same? If so, we can test a ludicrous n=23 task against a known number to see if properly finishes the calculation. It seems that n= 24 very quickly reaches a point of max err; but what if n=23 has the same problem, it just takes longer to show up?

Also, has anyone tested WR tasks on the new app? I did an n=21 task and it validated, but besides that, didn't really mess with the higher ranges.


I'm pretty sure the answer is yes, because the first thing Genefer does is expand b^N+1 into the actual number to be tested. So it shouldn't matter what N you use. But I thought I'd try it, just to make sure. The results are a bit troubling:

C:\Temp\GFN\OCL4>geneferocl4_windows-1004.exe -q "10^4096+1
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows-1004.exe -q 10^4096+1

Priority change succeeded.

Testing 10^4096+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1
CUDA' and driver '359.06'.


The checkpoint doesn't match current test: 10^4096+1 != 3066672^2048+1. Current
test will be restarted
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 10^4096+1 is 0:00:00

maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check
maxErr exceeded by all available transform implementations

C:\Temp\GFN\OCL4>geneferocl4_windows-1004.exe -q "10^8192+1
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows-1004.exe -q 10^8192+1

Priority change succeeded.

Testing 10^8192+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1
CUDA' and driver '359.06'.


The checkpoint doesn't match current test: 10^8192+1 != 10^4096+1. Current test
will be restarted
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 10^8192+1 is 0:00:00

maxErr exceeded for 10^8192+1, 1.0000 > 0.4500 during final check
maxErr exceeded by all available transform implementations


It simply fails.

I tried running -t, which automatically runs known primes, and it seemed to work:

C:\Temp\GFN\OCL4>geneferocl4_windows-1004.exe -t
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows-1004.exe -t

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 100014^32+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1
CUDA' and driver '359.06'.


The checkpoint doesn't match current test: 100014^32+1 != 10^8192+1. Current te
st will be restarted
Starting initialization...
Initialization complete (0.001 seconds).
100014^32+1 is a probable prime. (161 digits) (err = 0.0000) (time = 0:00:00) 16
:50:45

Testing 100234^64+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 100234^64+1 is 0:00:00
100234^64+1 is a probable prime. (321 digits) (err = 0.0000) (time = 0:00:00) 16
:50:45

Testing 100032^128+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 100032^128+1 is 0:00:00
100032^128+1 is a probable prime. (641 digits) (err = 0.0000) (time = 0:00:00) 1
6:50:45

Testing 5684328^256+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 5684328^256+1 is 0:00:00
5684328^256+1 is a probable prime. (1730 digits) (err = 0.0000) (time = 0:00:01)
16:50:46

Testing 4619000^512+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 4619000^512+1 is 0:00:00
4619000^512+1 is a probable prime. (3413 digits) (err = 0.0000) (time = 0:00:00)
16:50:46

Testing 3752220^1024+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 3752220^1024+1 is 0:00:00
3752220^1024+1 is a probable prime. (6733 digits) (err = 0.0000) (time = 0:00:01
) 16:50:47

Testing 3066672^2048+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 3066672^2048+1 is 0:00:01
3066672^2048+1 is a probable prime. (13285 digits) (err = 0.0000) (time = 0:00:0
1) 16:50:48

Testing 2485064^4096+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.004 seconds).
Estimated time remaining for 2485064^4096+1 is 0:00:02
2485064^4096+1 is a probable prime. (26196 digits) (err = 0.0000) (time = 0:00:0
3) 16:50:51

Testing 2030234^8192+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.005 seconds).
Estimated time remaining for 2030234^8192+1 is 0:00:06
2030234^8192+1 is a probable prime. (51672 digits) (err = 0.0000) (time = 0:00:0
6) 16:50:57


The -r residue test on composites works too:

C:\Temp\GFN\OCL4>geneferocl4_windows-1004.exe -r
geneferocl4 3.2.10-dev (Windows/OpenCL-NTT-32-2.3/32-bit)
Copyright 2001-2016, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2016, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: geneferocl4_windows-1004.exe -r

Priority change succeeded.
Running tests for transform implementation "OCL4"

Testing 10234^64+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1
CUDA' and driver '359.06'.


The checkpoint doesn't match current test: 10234^64+1 != 1651902^16384+1. Curre
nt test will be restarted
Starting initialization...
Initialization complete (0.001 seconds).
10234^64+1 is composite. (RES=e99e17786d3bacc2) (257 digits) (err = 0.0000) (tim
e = 0:00:00) 16:51:51

Testing 10032^128+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 10032^128+1 is 0:00:00
10032^128+1 is composite. (RES=1d9767ae8d652d65) (513 digits) (err = 0.0000) (ti
me = 0:00:00) 16:51:51

Testing 584328^256+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 584328^256+1 is 0:00:00
584328^256+1 is composite. (RES=32798be013527165) (1477 digits) (err = 0.0000) (
time = 0:00:01) 16:51:52

Testing 419000^512+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 419000^512+1 is 0:00:00
419000^512+1 is composite. (RES=fee1d2c0894a662f) (2879 digits) (err = 0.0000) (
time = 0:00:00) 16:51:52

Testing 352220^1024+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.001 seconds).
Estimated time remaining for 352220^1024+1 is 0:00:00
352220^1024+1 is composite. (RES=6c7ebc76febe0583) (5680 digits) (err = 0.0000)
(time = 0:00:01) 16:51:53

Testing 366672^2048+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 366672^2048+1 is 0:00:01
366672^2048+1 is composite. (RES=fd37a864102cf81e) (11396 digits) (err = 0.0000)
(time = 0:00:01) 16:51:54

Testing 285064^4096+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 285064^4096+1 is 0:00:02
285064^4096+1 is composite. (RES=f81ed37ab6df33f9) (22344 digits) (err = 0.0000)
(time = 0:00:03) 16:51:57

Testing 230234^8192+1...
Using OCL4 transform
Starting initialization...
Initialization complete (0.005 seconds).
Estimated time remaining for 230234^8192+1 is 0:00:05
230234^8192+1 is composite. (RES=5972a302c255e081) (43927 digits) (err = 0.0000)
(time = 0:00:05) 16:52:02


Note: The copy/paste above shows that the tests were run on v1004, however I re-ran the tests on 1007 (the latest version). 1007 produces identical results as 1004. All tests were done on a GTX 580, Windows 7 64 bit, driver 359.07

____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91553 - Posted: 25 Jan 2016 | 22:10:33 UTC - in response to Message 91546.

1) I see intermittent MaxErr errors using OCL. They don't happen very often, but they do occur frequently enough to be annoying on long tasks.

I tested more than 40 n=20 GFNs and no error occurred on my computer. All of them are validated.
The generated code depends on hardware and driver version: it would be interesting to note them because the bug could be in the compiler. OpenCL compilers are not reliable as C compilers.

4) My silly test of n=24 seems to fail for everyone who has tested it.

This one is repeatable then I will debug it.
But I don't see why it would be related to some "intermittent MaxErr".
Note that error detection was improved in OCL4 then an hardware error will probably generate a MaxErr=1 error rather than a wrong residue.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91555 - Posted: 25 Jan 2016 | 22:24:46 UTC - in response to Message 91553.

1) I see intermittent MaxErr errors using OCL. They don't happen very often, but they do occur frequently enough to be annoying on long tasks.

I tested more than 40 n=20 GFNs and no error occurred on my computer. All of them are validated.
The generated code depends on hardware and driver version: it would be interesting to note them because the bug could be in the compiler. OpenCL compilers are not reliable as C compilers.

4) My silly test of n=24 seems to fail for everyone who has tested it.

This one is repeatable then I will debug it.
But I don't see why it would be related to some "intermittent MaxErr".
Note that error detection was improved in OCL4 then an hardware error will probably generate a MaxErr=1 error rather than a wrong residue.


The two tests in my examples above (10^4096+1 and 10^8192+1) are repeatable and should be much easier to debug than my n=24 test!
____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91557 - Posted: 25 Jan 2016 | 22:28:13 UTC - in response to Message 91551.

1) I see intermittent MaxErr errors using OCL.

geneferocl4_windows-1004.exe -q 10^4096+1

Testing 10^4096+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1
CUDA' and driver '359.06'.

Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 10^4096+1 is 0:00:00

maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check

OK, this is not intermittent but repeatable.

Running on platform 'NVIDIA CUDA', device 'GeForce GT 740M', version 'OpenCL 1.2 CUDA' and driver '361.43'. Testing 10^4096+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 10^4096+1 is 0:00:00 maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check

and

Running on platform 'Intel(R) OpenCL', device 'Intel(R) HD Graphics 4000', version 'OpenCL 1.2 ' and driver '10.18.10.4276'. Testing 10^4096+1... Using OCL4 transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 10^4096+1 is 0:00:02 maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check

but

Running on platform 'NVIDIA CUDA', device 'GeForce GT 740M', version 'OpenCL 1.2 CUDA' and driver '361.43'. Testing 10^4096+1... Using OCL transform Starting initialization... Initialization complete (0.002 seconds). Estimated time remaining for 10^4096+1 is 0:00:00 10^4096+1 is composite. (RES=0804060302000106) (4097 digits) (err = 0.0000) (time = 0:00:02) 23:16:22


OCL4 is bugged :o(

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91560 - Posted: 25 Jan 2016 | 22:42:27 UTC - in response to Message 91548.

Quick question: if we get the same number written in for 2 different n, is the residue going to be the same? If so, we can test a ludicrous n=23 task against a known number to see if properly finishes the calculation. It seems that n= 24 very quickly reaches a point of max err; but what if n=23 has the same problem, it just takes longer to show up?


You know what, I was wrong about expecting the numbers to have the same residue. Ignoring the OCL4 problem, here's some results from genefer64 and genefercuda:

100^8192+1 is composite. (RES=561a094a410d4c46) (16385 digits) (err = 0.0000) (time = 0:00:02) 17:35:16 10^16384+1 is composite. (RES=0506030106070007) (16385 digits) (err = 0.0000) (time = 0:00:03) 17:35:36


Identical results with CUDA and FMA3. To my surprise, the residues are different.

(Interestingly, the CPU is faster than CUDA at those N's.)
____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91561 - Posted: 25 Jan 2016 | 22:50:55 UTC - in response to Message 91557.

Same bug with OCL3:

Testing 10^4096+1... Using OCL3 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 10^4096+1 is 0:00:01 maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91562 - Posted: 25 Jan 2016 | 23:02:39 UTC - in response to Message 91561.

Same bug with OCL3:

Testing 10^4096+1... Using OCL3 transform Starting initialization... Initialization complete (0.003 seconds). Estimated time remaining for 10^4096+1 is 0:00:01 maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check


OCL2 and OCL seem to work.

____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91563 - Posted: 25 Jan 2016 | 23:02:58 UTC - in response to Message 91551.
Last modified: 26 Jan 2016 | 0:00:29 UTC

It simply fails.


(EDIT: just removed the copyright messages on the logs. The post is big enough as it is.....)

(EDIT2: I've now tested N=4096 for every b up to 100. The only ones that MaxErr are b=4, 10 and 16. Now, because 16 is 4 squared, I decided to try b=16^2=256... and what a surprise, MaxErr again! Let's keep going with this power of 2 trend: b=256^2=65536, so in other words, 65536^4092+1 also MaxErr. Doing the same for base 10 only errors on 10; the others do fine.)

Ah... fails on me too! Card is Gtx 970, driver is 361.60 - very different test conditions. I tested b = 10 at different n values. n=5~11 (N=32~2048) and n=14 (N=16384) and above had no problem. But when it got to n=12 and n=13 (N=4096 and 8192), it MaxErr. Logs:

Command line: geneferocl4_1007_windows.exe -q 10^32+1

Priority change succeeded.

Testing 10^32+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
10^32+1 is composite. (RES=0306080608000007) (33 digits) (err = 0.0000) (time = 0:00:01) 20:39:05

--------------

Command line: geneferocl4_1007_windows.exe -q 10^64+1

Priority change succeeded.

Testing 10^64+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
10^64+1 is composite. (RES=0802030704030100) (65 digits) (err = 0.0000) (time = 0:00:00) 20:39:17

--------------

Command line: geneferocl4_1007_windows.exe -q 10^128+1

Priority change succeeded.

Testing 10^128+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.003 seconds).
10^128+1 is composite. (RES=0503020505010506) (129 digits) (err = 0.0000) (time = 0:00:00) 20:39:27

--------------

Command line: geneferocl4_1007_windows.exe -q 10^256+1

Priority change succeeded.

Testing 10^256+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
10^256+1 is composite. (RES=0407060106060509) (257 digits) (err = 0.0000) (time = 0:00:00) 20:39:36

--------------

Command line: geneferocl4_1007_windows.exe -q 10^512+1

Priority change succeeded.

Testing 10^512+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.003 seconds).
Estimated time remaining for 10^512+1 is 0:00:00
10^512+1 is composite. (RES=0504050809090308) (513 digits) (err = 0.0000) (time = 0:00:00) 20:39:43

--------------

Command line: geneferocl4_1007_windows.exe -q 10^1024+1

Priority change succeeded.

Testing 10^1024+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 10^1024+1 is 0:00:00
10^1024+1 is composite. (RES=0006000104060506) (1025 digits) (err = 0.0000) (time = 0:00:00) 20:39:49

--------------

Command line: geneferocl4_1007_windows.exe -q 10^2048+1

Priority change succeeded.

Testing 10^2048+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 10^2048+1 is 0:00:00
10^2048+1 is composite. (RES=0304080404090609) (2049 digits) (err = 0.0000) (time = 0:00:01) 20:39:58

--------------

Command line: geneferocl4_1007_windows.exe -q 10^4096+1

Priority change succeeded.

Testing 10^4096+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.005 seconds).
Estimated time remaining for 10^4096+1 is 0:00:00

maxErr exceeded for 10^4096+1, 1.0000 > 0.4500 during final check
maxErr exceeded by all available transform implementations


--------------

Command line: geneferocl4_1007_windows.exe -q 10^8192+1

Priority change succeeded.

Testing 10^8192+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Initialization complete (0.003 seconds).
Estimated time remaining for 10^8192+1 is 0:00:01

maxErr exceeded for 10^8192+1, 1.0000 > 0.4500 during final check
maxErr exceeded by all available transform implementations


--------------

Command line: geneferocl4_1007_windows.exe -q 10^16384+1

Priority change succeeded.

Testing 10^16384+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.003 seconds).
Estimated time remaining for 10^16384+1 is 0:00:02
10^16384+1 is composite. (RES=0506030106070007) (16385 digits) (err = 0.0000) (time = 0:00:03) 20:40:55

--------------

Command line: geneferocl4_1007_windows.exe -q 10^32768+1

Priority change succeeded.

Testing 10^32768+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.004 seconds).
Estimated time remaining for 10^32768+1 is 0:00:06
10^32768+1 is composite. (RES=0205000601040104) (32769 digits) (err = 0.0000) (time = 0:00:07) 20:41:16

--------------

Command line: geneferocl4_1007_windows.exe -q 10^65536+1

Priority change succeeded.

Testing 10^65536+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.006 seconds).
Estimated time remaining for 10^65536+1 is 0:00:22
10^65536+1 is composite. (RES=0804000400030709) (65537 digits) (err = 0.0000) (time = 0:00:20) 20:47:14



But that's not all. Oh it's just the start of it. Since only 4096 and 8192 showed any problem, I decided to test it further. Let's change B; instead of 10, how about 100? And guess what: it runs with b=100
Command line: geneferocl4_1007_windows.exe -q 100^4096+1

Priority change succeeded.

Testing 100^4096+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.003 seconds).
Estimated time remaining for 100^4096+1 is 0:00:01
100^4096+1 is composite. (RES=212d5e461a221729) (8193 digits) (err = 0.0000) (time = 0:00:02) 20:51:25

--------------

Command line: geneferocl4_1007_windows.exe -q 100^8192+1

Priority change succeeded.

Testing 100^8192+1...
Using OCL4 transform

Running on platform 'NVIDIA CUDA', device 'GeForce GTX 970', version 'OpenCL 1.2 CUDA' and driver '361.60'.

Starting initialization...
Initialization complete (0.002 seconds).
Estimated time remaining for 100^8192+1 is 0:00:02
100^8192+1 is composite. (RES=561a094a410d4c46) (16385 digits) (err = 0.0000) (time = 0:00:03) 20:54:00


Now, wait a second..... 100^8192+1 should be the same as 10^16384+1 (or do am i bad at math and got it wrong?), so let's look at the residues:

N=8192 - RES=561a094a410d4c46

N=16384 - RES=0506030106070007


I'm.... speechless. All my testing in Boinc showed no problem at all; which in a sense, it's a good thing. Still, it looks like there are still many problems with the OCL4 app. Should I keep randomly testing stuff or wait a newer version?

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91573 - Posted: 26 Jan 2016 | 9:35:55 UTC - in response to Message 91563.

Now, wait a second..... 100^8192+1 should be the same as 10^16384+1 (or do am i bad at math and got it wrong?), so let's look at the residues:

N=8192 - RES=561a094a410d4c46

N=16384 - RES=0506030106070007

I'm.... speechless. All my testing in Boinc showed no problem at all; which in a sense, it's a good thing. Still, it looks like there are still many problems with the OCL4 app. Should I keep randomly testing stuff or wait a newer version?

Don't see problems when the result is correct.

10^16384+1 => base is 10.
RES = 05 06 03 01 06 07 00 07
100^8192+1 => base is 100:
RES = 0x56 0x1a 0x09 0x4a 0x41 0x0d 0x4c 0x46 = 86 26 09 74 65 13 76 70.

Now, read the result in reverse order (from right to left):
07 00 07 06 01 03 06 05 <=> 70 76 13 65

I found "Mike's bug". It can occur if N/b is "large" (I will define "large" later). Then this is not a problem with current Boinc search.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91576 - Posted: 26 Jan 2016 | 10:58:02 UTC - in response to Message 91573.

10^4096+1 and 10^8192+1 bug with OCL3/OCL4 can occur if N > b^3 then not in Boinc ranges.

28224534^16777216+1 bug is another problem. I'm working on it.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91582 - Posted: 26 Jan 2016 | 15:40:52 UTC - in response to Message 91576.
Last modified: 26 Jan 2016 | 16:13:14 UTC

Exponent 16777216 (or larger) cannot be tested with OCL4 because the NTT uses the prime P = 127 * 2^24 + 1.
A root of order 2N = 2^25 doesn't exist modulo P and the algorithm fails.

400000000^8388608+1 seems to be OK:

Testing 400000000^8388608+1... Using OCL4 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 680', version 'OpenCL 1.2 CUDA' and driver '361.43'. Starting initialization... Initialization complete (351.770 seconds). Estimated time remaining for 400000000^8388608+1 is 1652:48:09 Testing 400000000^8388608+1... 239681536 steps to go (1647:43:00 remaining)


Any b^16777216+1 fails.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91583 - Posted: 26 Jan 2016 | 16:11:55 UTC

geneferocl4_windows.exe RC2 is committed (1019): fix 10^4096+1 and 10^8192+1 bug.

HonzaProject donor
Volunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1893
ID: 352
Credit: 3,141,484,578
RAC: 40,664
Discovered 7 mega primesEliminated 4 conjecture "k"sFound 2 primes in the 2018 Tour de PrimesFound 1 prime in the 2018 Tour de Primes Mountain Stage2019 Tour de Primes largest primeFound 4 primes in the 2019 Tour de PrimesFound 1 mega prime in the 2019 Tour de PrimesFound 1 prime in the 2019 Tour de Primes Mountain StageFound 1 prime in the 2020 Tour de Primes321 LLR Double Bronze: Earned 100,000,000 credits (100,425,701)Cullen LLR Emerald: Earned 50,000,000 credits (60,526,018)ESP LLR Emerald: Earned 50,000,000 credits (50,853,190)Generalized Cullen/Woodall LLR Emerald: Earned 50,000,000 credits (50,309,119)PPS LLR Double Bronze: Earned 100,000,000 credits (122,476,468)PSP LLR Emerald: Earned 50,000,000 credits (50,563,867)SoB LLR Double Bronze: Earned 100,000,000 credits (101,573,710)SR5 LLR Double Bronze: Earned 100,000,000 credits (101,883,529)SGS LLR Emerald: Earned 50,000,000 credits (55,573,529)TPS LLR (retired) Bronze: Earned 10,000 credits (43,033)TRP LLR Double Bronze: Earned 100,000,000 credits (100,338,924)Woodall LLR Emerald: Earned 50,000,000 credits (50,956,094)321 Sieve Double Bronze: Earned 100,000,000 credits (115,948,450)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,142,109)Generalized Cullen/Woodall Sieve (suspended) Emerald: Earned 50,000,000 credits (50,504,945)PPS Sieve Double Gold: Earned 500,000,000 credits (513,057,580)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,288,222)TRP Sieve (suspended) Sapphire: Earned 20,000,000 credits (20,149,354)AP 26/27 Double Silver: Earned 200,000,000 credits (232,724,773)GFN Double Gold: Earned 500,000,000 credits (764,816,358)PSA Double Gold: Earned 500,000,000 credits (535,232,795)
Message 91584 - Posted: 26 Jan 2016 | 16:42:38 UTC
Last modified: 26 Jan 2016 | 16:47:59 UTC

OCL4 V1019 seems to work on AMD 280X.
EDIT: And on Fury Nano as well.
____________
My stats
Badge score: 1*1 + 5*1 + 8*3 + 9*11 + 10*1 + 11*1 + 12*3 = 186

Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 91586 - Posted: 26 Jan 2016 | 17:30:16 UTC

OCL4 for Linux32/64 (Fix OCL3/OCL4 10^4096+1 and 10^8192+1 bug) is available at https://www.assembla.com/spaces/genefer/subversion/commits/1020.
____________
Best wishes. Knowledge is power. by jjwhalen

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91587 - Posted: 26 Jan 2016 | 17:51:51 UTC - in response to Message 91582.
Last modified: 26 Jan 2016 | 18:18:16 UTC

EDIT: Tests were done with v1007. Since I already wrote all of this, I'll just leave it as is, re-test everything again, and make sure it's all fixed.


EDIT2: I've now tested v1019. The 2^(2^k) problem is still there, but at least the random failing B are out. I still pledge for an n=24 built in error code, though....

Any b^16777216 fails.

If it's known to fail for higher N, might as well include a "N is too high, program can't run" message, just like you include a "b is odd so number is even" message. I bet

Anyway, after more testing, I decided to compile a list of the known failures. There seems to be 2 rules for this:
1- 2^ (2^k) always fails. For some N, this includes k=0, for others it's only k>0. So that's b=2, 4, 16, 256 and 65536.
2- For "small" b (possibly N>b^3 ?), then there MIGHT be a bug for some of the candidates.

So here's a compiled list for easy reading:
N=1024 fails for k>=0.
N=2048 fails for k>=0. No other failures.
N=4096 fails for k>0. Also fails for b=10
N=8192 fails for k>=0. Also fails for b=10 and b=12.
N=16384 fails for k>=0. Also fails for b=12 and b=14.
N=32768 fails for k>=0. Also fails for b=12, b=14, b=18 and b=20
N=65536 fails for k>=0. Also fails for b=12, b=14, b=18, b=20, b=22, b=24, b=26 and b=32.
N=131072 fails for k>=0. Also fails for b=14, b=18, b=20, b=22, b=24, b=26, b=28, b=30, b=32 and b=34.


While most of this is outside of our Boinc test realm, it's important to note that n=21 and n=22 will eventually fail once we get to b=65536 (assuming that such candidate would be sent)... and in fact, leading edge for WR is 64330, so if that candidate is still there, we should remove it now.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91588 - Posted: 26 Jan 2016 | 18:37:13 UTC - in response to Message 91587.

While most of this is outside of our Boinc test realm, it's important to note that n=21 and n=22 will eventually fail once we get to b=65536 (assuming that such candidate would be sent)... and in fact, leading edge for WR is 64330, so if that candidate is still there, we should remove it now.


b=65536 is not in the sieve file for either n=21 or n=22. Or at least that appears to be the case. There's some irregularity with n=22 that needs to be resolved. N=21 is definite. N=22 will take some more research.
____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91589 - Posted: 26 Jan 2016 | 18:55:58 UTC - in response to Message 91588.
Last modified: 26 Jan 2016 | 19:32:24 UTC

While most of this is outside of our Boinc test realm, it's important to note that n=21 and n=22 will eventually fail once we get to b=65536 (assuming that such candidate would be sent)... and in fact, leading edge for WR is 64330, so if that candidate is still there, we should remove it now.


b=65536 is not in the sieve file for either n=21 or n=22. Or at least that appears to be the case. There's some irregularity with n=22 that needs to be resolved. N=21 is definite. N=22 will take some more research.

Unless it get's fixed, if someone using the OCL4 (and OCL3) app happens to get such task, their task WILL fail. So that's why I'm suggesting to remove that particular candidate and separately test it outside of Boinc, as people using the app_info might end up getting a known-to-be-bad task.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91590 - Posted: 26 Jan 2016 | 19:17:26 UTC - in response to Message 91587.

1- 2^ (2^k) always fails. For some N, this includes k=0, for others it's only k>0. So that's b=2, 4, 16, 256 and 65536.

OCL3 fails too. I will fix it. Again a false error detection.

Profile JeppeSNProject donor
Avatar
Send message
Joined: 5 Apr 14
Posts: 1378
ID: 306875
Credit: 21,623,822
RAC: 2,050
Found 1 prime in the 2020 Tour de Primes321 LLR Gold: Earned 500,000 credits (529,293)Cullen LLR Gold: Earned 500,000 credits (611,298)ESP LLR Silver: Earned 100,000 credits (139,922)Generalized Cullen/Woodall LLR Bronze: Earned 10,000 credits (35,236)PPS LLR Turquoise: Earned 5,000,000 credits (9,594,179)PSP LLR Silver: Earned 100,000 credits (212,242)SoB LLR Silver: Earned 100,000 credits (237,390)SR5 LLR Silver: Earned 100,000 credits (145,419)SGS LLR Silver: Earned 100,000 credits (105,212)TRP LLR Silver: Earned 100,000 credits (342,501)Woodall LLR Silver: Earned 100,000 credits (109,455)321 Sieve Silver: Earned 100,000 credits (175,037)PPS Sieve Bronze: Earned 10,000 credits (10,113)AP 26/27 Bronze: Earned 10,000 credits (12,129)GFN Amethyst: Earned 1,000,000 credits (1,674,106)PSA Turquoise: Earned 5,000,000 credits (7,614,290)
Message 91594 - Posted: 26 Jan 2016 | 20:08:17 UTC - in response to Message 91587.

Rafael, 65536^(2^n) = (2^(2^4))^(2^n) = 2^(2^(4+n)) = F (4+n) is a classical Fermat number. It will be skipped because it has a known factor. /JeppeSN

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91598 - Posted: 26 Jan 2016 | 21:11:46 UTC - in response to Message 91588.

b=65536 is not in the sieve file for either n=21 or n=22. Or at least that appears to be the case. There's some irregularity with n=22 that needs to be resolved. N=21 is definite. N=22 will take some more research.


65536 (and similar numbers) are definitely removed from those sieve files, so there's no chance of the server sending them out for testing. I just needed to double check what I believed to be true.
____________
My lucky number is 75898524288+1

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91610 - Posted: 27 Jan 2016 | 9:59:24 UTC

geneferocl4_windows.exe RC3 is committed (1022): fix Fermat numbers test and n>=24 -> print error.

Profile rroonnaallddProject donor
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
321 LLR Silver: Earned 100,000 credits (101,692)Cullen LLR Silver: Earned 100,000 credits (104,876)ESP LLR Silver: Earned 100,000 credits (101,979)PPS LLR Silver: Earned 100,000 credits (148,018)PSP LLR Silver: Earned 100,000 credits (140,441)SoB LLR Silver: Earned 100,000 credits (119,475)SR5 LLR Silver: Earned 100,000 credits (120,939)SGS LLR Silver: Earned 100,000 credits (122,783)TRP LLR Silver: Earned 100,000 credits (100,115)Woodall LLR Silver: Earned 100,000 credits (107,459)321 Sieve Silver: Earned 100,000 credits (202,757)Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (6,908,135)PPS Sieve Sapphire: Earned 20,000,000 credits (25,450,104)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Silver: Earned 100,000 credits (130,966)TRP Sieve (suspended) Silver: Earned 100,000 credits (201,525)AP 26/27 Silver: Earned 100,000 credits (100,015)GFN Silver: Earned 100,000 credits (246,369)PSA Silver: Earned 100,000 credits (226,594)
Message 91612 - Posted: 27 Jan 2016 | 15:09:06 UTC - in response to Message 91610.

geneferocl4_windows.exe RC3 is committed (1022): fix Fermat numbers test and n>=24 -> print error.

Same app for Linux32/64 is available at https://www.assembla.com/spaces/genefer/subversion/commits/1025
____________
Best wishes. Knowledge is power. by jjwhalen

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91614 - Posted: 27 Jan 2016 | 17:54:03 UTC - in response to Message 91610.

geneferocl4_windows.exe RC3 is committed (1022): fix Fermat numbers test and n>=24 -> print error.

Testing b=2, N=64~65536 says it's a probable prime..... then you get to n=17 and above, and it beautifully MaxErr there. Only happens with with b=2, though, the other numbers all did fine.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91625 - Posted: 27 Jan 2016 | 23:18:59 UTC - in response to Message 91614.

Testing b=2, N=64~65536 says it's a probable prime..... then you get to n=17 and above, and it beautifully MaxErr there. Only happens with with b=2, though, the other numbers all did fine.

I leave it at that. I prefer a false error detection with is composite number (as anybody knows for over 30 years) to a true computation error that would not be detected with another GFN.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91626 - Posted: 27 Jan 2016 | 23:27:42 UTC - in response to Message 91625.

Testing b=2, N=64~65536 says it's a probable prime..... then you get to n=17 and above, and it beautifully MaxErr there. Only happens with with b=2, though, the other numbers all did fine.

I leave it at that. I prefer a false error detection with is composite number (as anybody knows for over 30 years) to a true computation error that would not be detected with another GFN.

Okay, so 1022 is looking like the "final" version, assuming nothing else is found? If so, then I guess, I'll start my last debug section by trying it on my AMD and Intel iGPUs. Just to be sure....

(Oh, btw, it still won't work on the my Ivy Bridge mobile iGPU, though I've long given up on that dream. It wouldn't be a useful GPU anyways.....)

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91629 - Posted: 28 Jan 2016 | 1:12:53 UTC
Last modified: 28 Jan 2016 | 1:53:38 UTC

Right... so I'm testing my APU and the Skylake iGPU (so far, so good) and I'm encountering a bit of a problem. I know it'll sound a bit vague, but the Skylake iGPU is freezing sometimes. For the majority of the tests, it's perfectly normal. But once in a while, the program will just freeze (and this seems to happen mostly on n=14 and n=15, in my n=6~15 batch of tests).

It would always complete the initialization, give a few time estimates, and then it just stops, without ever completing the test. You can see that the program is doing nothing, as the iGPU clock goes down to 0 and never up. It just stays there saying "testing, xx steps to go (xx remaining)" endlessly. Unlike my Ivy Bridge iGPU (which never even initalizes, btw), using the Ctrl+c command on Windows does say "^C caught. Writing checkpoint", but it never actually finishes writing it, it just stays there staring at you forever.

It's not a problem with a particular number: once I find a number where it happens, I try it multiple times and it never happens again. Then all of a sudden, I'm doing a different test and it just happens. Not once on my APU nor my GTX, but I've already seen it at least 4 different times on the Skylake iGPU.

Good luck diagnosing that......... (then again, it's an Intel iGPU, so I don't know if it's worth the effort).


EDIT: Because I couldn't reliably replicate it, I decided to put a camera and record my screen, to have evidence for when it would eventually happen again. And it did. At around 1:30 there's a second of hiccup, probably the checkpoint moment (happens just before printing a new estimate). Then at around 2:05, it happens. When the mouse shows up, that's when I realized it happened.

As soon as the upload finishes, you can see it here: https://youtu.be/sYO434jK99E

Profile JeppeSNProject donor
Avatar
Send message
Joined: 5 Apr 14
Posts: 1378
ID: 306875
Credit: 21,623,822
RAC: 2,050
Found 1 prime in the 2020 Tour de Primes321 LLR Gold: Earned 500,000 credits (529,293)Cullen LLR Gold: Earned 500,000 credits (611,298)ESP LLR Silver: Earned 100,000 credits (139,922)Generalized Cullen/Woodall LLR Bronze: Earned 10,000 credits (35,236)PPS LLR Turquoise: Earned 5,000,000 credits (9,594,179)PSP LLR Silver: Earned 100,000 credits (212,242)SoB LLR Silver: Earned 100,000 credits (237,390)SR5 LLR Silver: Earned 100,000 credits (145,419)SGS LLR Silver: Earned 100,000 credits (105,212)TRP LLR Silver: Earned 100,000 credits (342,501)Woodall LLR Silver: Earned 100,000 credits (109,455)321 Sieve Silver: Earned 100,000 credits (175,037)PPS Sieve Bronze: Earned 10,000 credits (10,113)AP 26/27 Bronze: Earned 10,000 credits (12,129)GFN Amethyst: Earned 1,000,000 credits (1,674,106)PSA Turquoise: Earned 5,000,000 credits (7,614,290)
Message 91646 - Posted: 28 Jan 2016 | 6:41:54 UTC - in response to Message 91629.

Right... so I'm testing my APU and the Skylake iGPU (so far, so good) and I'm encountering a bit of a problem. I know it'll sound a bit vague, but the Skylake iGPU is freezing sometimes. For the majority of the tests, it's perfectly normal. But once in a while, the program will just freeze


It sounds like the same Skylake bug mentioned in Upcoming BIOS patch for Skylake CPUs. Or wait, your problem is with Skylake GPU, and the known problem with Skylake CPU? /JeppeSN

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91649 - Posted: 28 Jan 2016 | 9:20:29 UTC - in response to Message 91629.

Right... so I'm testing my APU and the Skylake iGPU (so far, so good) and I'm encountering a bit of a problem. I know it'll sound a bit vague, but the Skylake iGPU is freezing sometimes.

I checked OCL4 on Intel HD Graphics 4000 and 4600 and I didn't have any problem.
I'm running residue test on HD Graphics 4600 (i7-4790) ... so far, so good.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91657 - Posted: 28 Jan 2016 | 13:31:08 UTC - in response to Message 91649.

I checked a megaprime on HD Graphics 4600:

Testing 42654182^131072+1... Using OCL4 transform Running on platform 'Intel(R) OpenCL', device 'Intel(R) HD Graphics 4600', version 'OpenCL 1.2 ' and driver '10.18.15.4256'. Starting initialization... Initialization complete (0.527 seconds). Estimated time remaining for 42654182^131072+1 is 3:34:44 42654182^131072+1 is a probable prime. (1000075 digits) (err = 0.0000) (time = 3:36:03) 14:10:57

Rafael, I noticed that your driver is '20.19.15.4331' ('OpenCL 2.0 '). It could be the reason for the instability.

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91660 - Posted: 28 Jan 2016 | 14:47:39 UTC - in response to Message 91657.
Last modified: 28 Jan 2016 | 14:48:20 UTC

Rafael, I noticed that your driver is '20.19.15.4331' ('OpenCL 2.0 '). It could be the reason for the instability.

Uh..... Intel? Are you okay?
http://imgur.com/LenEdPU

Anyway, I downloaded the first one. Strangely enough, the driver is showing 20.19.15.4352 in device manager (and obviously in the GFN program as well) instead of 15.10.14.4352 listed in Intel's site. Will conduct more testing on it and see if it's solved. So far, still no freezes....


I checked OCL4 on Intel HD Graphics 4000 and 4600 and I didn't have any problem.
I'm running residue test on HD Graphics 4600 (i7-4790) ... so far, so good.

The thing is, in the instances where it didn't freeze, the calculation worked flawlessly for eveyr number I tried. And even when it did freeze, closing the program and restarting from last checkpoint was no problem either: the checkpoint was detected, the program resumed from there, and the final results were always as expected.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91664 - Posted: 28 Jan 2016 | 16:24:41 UTC

I've got something unusual, consistent, and repeatable. I'm not sure what's happening here.

I'm trying to decide which version of OCL to run for GFN-16. Choices are OCL3 or OCL4. Ignore the "hogs a CPU core" problem. I'm just looking at which runs the tasks fastest.

Here's the benchmarks:

Command line: geneferocl3 -b Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'. 1203210^65536+1 Time: 157 us/mul. Err: 0.0000 398482 digits Genefer Mark = 39.


Command line: geneferocl4_windows-1022.exe -b Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'. 1203210^65536+1 Time: 137 us/mul. Err: 0.0000 398482 digits Genefer Mark = 52.


Going by the benchmarks, OCL4 is clearly faster.

But look at actual calculations:

Command line: geneferocl3 -q 6000000^65536+1 Using OCL3 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'. Starting initialization... Initialization complete (0.123 seconds). Estimated time remaining for 6000000^65536+1 is 0:03:55


Command line: geneferocl4_windows-1022.exe -q 6000000^65536+1 Using OCL4 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'. Starting initialization... Initialization complete (0.089 seconds). Estimated time remaining for 6000000^65536+1 is 0:04:44


In the benchmarks, OCL4 is faster. In the calculation estimates, OCL3 is faster.

The benchmarks should be using the same timing mechanism as the runtime estimates, so there should not be a discrepancy between the benchmarks and the runtimes, unless changing the b value is changing the iteration times. I'm pretty sure that the iteration times used to be constant for a given N. Has that changed?
____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91666 - Posted: 28 Jan 2016 | 16:43:36 UTC - in response to Message 91664.
Last modified: 28 Jan 2016 | 16:49:18 UTC

In the benchmarks, OCL4 is faster. In the calculation estimates, OCL3 is faster.

The benchmarks should be using the same timing mechanism as the runtime estimates, so there should not be a discrepancy between the benchmarks and the runtimes, unless changing the b value is changing the iteration times. I'm pretty sure that the iteration times used to be constant for a given N. Has that changed?

Look at the B limits for when OCL4 switches to the 3 prime (slower) transform. 6M is past the 5.7M point, so that's why it's so much slower.


EDIT: Yves, it freezes with the supposedly 15.10.14.4352. Will try the others.

Yves Gallot
Volunteer developer
Project scientist
Send message
Joined: 19 Aug 12
Posts: 644
ID: 164101
Credit: 305,010,093
RAC: 78
GFN Double Silver: Earned 200,000,000 credits (305,010,093)
Message 91667 - Posted: 28 Jan 2016 | 16:54:43 UTC - in response to Message 91666.
Last modified: 28 Jan 2016 | 17:10:10 UTC

In the benchmarks, OCL4 is faster. In the calculation estimates, OCL3 is faster.

Look at the B limits for when OCL4 switches to the 3 prime (slower) transform. 6M is past the 5.7M point, so that's why it's so much slower.

Yes, OCL4 prints 2 benchmarks. On my computer:

1203210^65536+1 Time: 102 us/mul. Err: 0.0000 398482 digits Genefer Mark = 66. 100000000^65536+1 Time: 169 us/mul. Err: 0.0000 524289 digits Genefer Mark = 39.


EDIT: Rafael, I updated the computer and I'm running the test with driver '20.19.15.4331'. But OpenCL 1.2 compiler is used on a HD Graphics 4600 (not compatible with OpenCL 2.0).

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)Generalized Cullen/Woodall Sieve (suspended) Jade: Earned 10,000,000 credits (10,037,204)PPS Sieve Jade: Earned 10,000,000 credits (10,305,147)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Ruby: Earned 2,000,000 credits (2,000,053)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,030,160)AP 26/27 Jade: Earned 10,000,000 credits (12,282,634)GFN Sapphire: Earned 20,000,000 credits (29,911,497)PSA Double Bronze: Earned 100,000,000 credits (170,761,999)
Message 91671 - Posted: 28 Jan 2016 | 17:50:38 UTC - in response to Message 91667.
Last modified: 28 Jan 2016 | 18:30:01 UTC

EDIT: Rafael, I updated the computer and I'm running the test with driver '20.19.15.4331'. But OpenCL 1.2 compiler is used on a HD Graphics 4600 (not compatible with OpenCL 2.0).

And I am finally using 10.18.15.4279. Intel actually says that's 15.40.7.4279, but device manager and boinc are telling me 10.18.15.4279. At any rate, the other drivers all failed on me, except maybe this one, as I just started testing. EDIT: nope, didn't work either.

Oh, and a question. Shouldn't the original OCL app work, since the device is OpenCL 2.0 rated? When I run it, it jumps straight to the discrete GPU instead of letting me choose the iGPU if I want to.

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91672 - Posted: 28 Jan 2016 | 17:52:32 UTC - in response to Message 91666.

In the benchmarks, OCL4 is faster. In the calculation estimates, OCL3 is faster.

The benchmarks should be using the same timing mechanism as the runtime estimates, so there should not be a discrepancy between the benchmarks and the runtimes, unless changing the b value is changing the iteration times. I'm pretty sure that the iteration times used to be constant for a given N. Has that changed?

Look at the B limits for when OCL4 switches to the 3 prime (slower) transform. 6M is past the 5.7M point, so that's why it's so much slower.


Oh, of course you're right. I was thinking of the older (v955) version of OCL4 that had a b limit of about 9M.

Command line: geneferocl4_windows-955.exe -q 6000000^65536+1 Using OCL4 transform Running on platform 'NVIDIA CUDA', device 'GeForce GTX 580', version 'OpenCL 1.1 CUDA' and driver '359.06'. Estimated time remaining for 6000000^65536+1 is 0:03:10


That's still the fastest in that range, and is in fact faster than OCL3.

I've asked Iain if the combined app can include the 955 version of the transform, as well as the two transforms from the current OCL4. Yves, is there any reason why that's not advisable?
____________
My lucky number is 75898524288+1

Profile Michael GoetzProject donor
Volunteer moderator
Project administrator
Avatar
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 236,922,854
RAC: 3,199
Found 5 primes in the 2021 Tour de PrimesFound 5 mega primes in the 2021 Tour de PrimesThe "Shut up already!" badge:  This loud mouth has mansplained on the forums over 10 thousand times!  Sheesh!!!Discovered the World's First GFN-19 prime!!!Discovered 6 mega primesFound 1 prime in the 2018 Tour de PrimesFound 1 prime in the 2019 Tour de PrimesFound 1 prime in the 2020 Tour de Primes321 LLR Ruby: Earned 2,000,000 credits (2,822,730)Cullen LLR Ruby: Earned 2,000,000 credits (3,624,591)ESP LLR Turquoise: Earned 5,000,000 credits (5,021,269)Generalized Cullen/Woodall LLR Ruby: Earned 2,000,000 credits (2,145,754)PPS LLR Jade: Earned 10,000,000 credits (16,008,485)PSP LLR Turquoise: Earned 5,000,000 credits (5,197,957)SoB LLR Sapphire: Earned 20,000,000 credits (34,291,181)SR5 LLR Jade: Earned 10,000,000 credits (10,007,110)SGS LLR Ruby: Earned 2,000,000 credits (3,252,256)TRP LLR Turquoise: Earned 5,000,000 credits (5,084,329)Woodall LLR Ruby: Earned 2,000,000 credits (2,911,985)321 Sieve Jade: Earned 10,000,000 credits (10,061,196)Cullen/Woodall Sieve (suspended) Ruby: Earned 2,000,000 credits (4,170,256)Generalized Cullen/Woodall Sieve (suspended) Turquoise: Earned 5,000,000 credits (5,059,304)PPS Sieve Sapphire: Earned 20,000,000 credits (22,885,121)Sierpinski (ESP/PSP/SoB) Sieve (suspended) Amethyst: Earned 1,000,000 credits (1,035,522)TRP Sieve (suspended) Ruby: Earned 2,000,000 credits (2,051,121)AP 26/27 Jade: Earned 10,000,000 credits (10,118,303)GFN Emerald: Earned 50,000,000 credits (76,733,355)PSA Jade: Earned 10,000,000 credits (12,445,029)
Message 91673 - Posted: 28 Jan 2016 | 17:54:12 UTC - in response to Message 91671.

EDIT: Rafael, I updated the computer and I'm running the test with driver '20.19.15.4331'. But OpenCL 1.2 compiler is used on a HD Graphics 4600 (not compatible with OpenCL 2.0).

And I am finally using 10.18.15.4279. Intel actually says that's 15.40.7.4279, but device manager and boinc are telling me 10.18.15.4279. At any rate, the other drivers all failed on me, except maybe this one, as I just started testing. I'll let it run for a while.

Oh, and a question. Shouldn't the original OCL app work, since the device is OpenCL 2.0 rated? When I run it, it jumps straight to the discrete GPU instead of letting me choose the iGPU if I want to.


OCL requires double precision. Does the iGPU have double precision hardware?
____________
My lucky number is 75898524288+1

Profile Rafael
Volunteer tester
Avatar
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 3,106
321 LLR Turquoise: Earned 5,000,000 credits (8,236,942)Cullen LLR Turquoise: Earned 5,000,000 credits (8,028,695)ESP LLR Turquoise: Earned 5,000,000 credits (8,027,771)Generalized Cullen/Woodall LLR Turquoise: Earned 5,000,000 credits (7,259,087)PPS LLR Turquoise: Earned 5,000,000 credits (7,722,618)PSP LLR Turquoise: Earned 5,000,000 credits (7,137,796)SoB LLR Turquoise: Earned 5,000,000 credits (6,941,728)SR5 LLR Turquoise: Earned 5,000,000 credits (7,186,115)SGS LLR Turquoise: Earned 5,000,000 credits (7,263,666)TRP LLR Turquoise: Earned 5,000,000 credits (8,751,781)Woodall LLR Turquoise: Earned 5,000,000 credits (7,119,125)321 Sieve Jade: Earned 10,000,000 credits (10,033,828)