1)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
New Version Testing
(Message 147201)
Posted 887 days ago by McDaWisel
Tests on a single threaded Mac ran clean as well:
MacBook Pro i9-9880H:
18446744013709551615 18446744073709551615
509DE6503B2A6E8E
1 6e10
5255 special instances
2 primes
8B77D2DA053842DD
227630e10 227636e10
2276306935816523 is a Wieferich special instance (-1 -3 p)
6522CBC4CA2E4CFF
338772621946054253 338772681946054253
338772621946054253 is a WallSunSun special instance (+0 +91 p)
589A69274A6AB5A8
|
2)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
New Version Testing
(Message 147183)
Posted 887 days ago by McDaWisel
I am running the single thread CPU tests now on the same MacBook Pro.
Not sure if you need this info but it's a i9-9880H.
|
3)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
New Version Testing
(Message 147182)
Posted 887 days ago by McDaWisel
Just ran them again.
Late 2019 MacBook Pro
AMD Radeon Pro 5500M Compute Engine
All clear.
18446744013709551615 18446744073709551615:
509DE6503B2A6E8E
1 6e10:
5255 instances + 2 primes
8B77D2DA053842DD
227630e10 227636e10:
2276306935816523 is a Wieferich special instance (-1 -3 p)
6522CBC4CA2E4CFF
338772621946054253 338772681946054253:
338772621946054253 is a WallSunSun special instance (+0 +91 p)
589A69274A6AB5A8
|
4)
Message boards :
Wieferich and Wall-Sun-Sun Prime Search :
New Version Testing
(Message 147176)
Posted 887 days ago by McDaWisel
I ran the GPU tests on a MacBook Pro, late 2019 using an AMD Radeon 5500M, here the results from the stderr.txt, let me know if you need something more:
shmget in attach_shmem: Invalid argument
19:37:54 (18621): Can't set up shared mem: -1. Will run in standalone mode.
WWocl version 1.03 by Bryan Little and Yves Gallot
Compiled Dec 29 2020 with GCC Apple LLVM 12.0.0 (clang-1200.0.32.28)
A Wieferich and WallSunSun prime number search program
Based on wwwwcl by Mark Rodenkirch and primesieve by Kim Walisch
Command line: ./WW_opencl_macintel64 18446744013709551615 18446744073709551615
Starting search at: 18446744013709551615
Stopping search at: 18446744073709551615
GPU Info:
Name: AMD Radeon Pro 5500M Compute Engine
Vendor: AMD
Driver: 1.2 (Oct 29 2020 23:01:05)
Compute Units: 24
Kernel profile: 12.705 ms. Estimated / Actual worksize: 26400000 / 22857022
Verifying GPU kernels using range 18446744013709551615 to 18446744013711551615...
GPU prime generator kernel test passed. (44952 PRPs)
GPU Wieferich/WallSunSun kernel tests passed. (193750445344149 checksum, 0 results)
Starting search...
Workunit complete. Average speed: 16.61 Million primes per second.
19:39:26 (18621): called boinc_finish(0)
shmget in attach_shmem: Invalid argument
19:40:17 (18669): Can't set up shared mem: -1. Will run in standalone mode.
WWocl version 1.03 by Bryan Little and Yves Gallot
Compiled Dec 29 2020 with GCC Apple LLVM 12.0.0 (clang-1200.0.32.28)
A Wieferich and WallSunSun prime number search program
Based on wwwwcl by Mark Rodenkirch and primesieve by Kim Walisch
Command line: ./WW_opencl_macintel64 1 6e10
START below minimum. Starting search at 127
Starting search at: 127
Stopping search at: 60000000000
GPU Info:
Name: AMD Radeon Pro 5500M Compute Engine
Vendor: AMD
Driver: 1.2 (Oct 29 2020 23:01:05)
Compute Units: 24
Kernel profile: 6.692 ms. Estimated / Actual worksize: 26400000 / 43395790
Verifying GPU kernels using range 127 to 2000127...
GPU prime generator kernel test passed. (148957 PRPs)
GPU Wieferich/WallSunSun kernel tests passed. (142810567078 checksum, 3114 results)
Starting search...
Workunit complete. Average speed: 39.47 Million primes per second.
19:41:24 (18669): called boinc_finish(0)
shmget in attach_shmem: Invalid argument
19:42:15 (18697): Can't set up shared mem: -1. Will run in standalone mode.
WWocl version 1.03 by Bryan Little and Yves Gallot
Compiled Dec 29 2020 with GCC Apple LLVM 12.0.0 (clang-1200.0.32.28)
A Wieferich and WallSunSun prime number search program
Based on wwwwcl by Mark Rodenkirch and primesieve by Kim Walisch
Command line: ./WW_opencl_macintel64 227630e10 227636e10
Starting search at: 2276300000000000
Stopping search at: 2276360000000000
GPU Info:
Name: AMD Radeon Pro 5500M Compute Engine
Vendor: AMD
Driver: 1.2 (Oct 29 2020 23:01:05)
Compute Units: 24
Kernel profile: 10.747 ms. Estimated / Actual worksize: 26400000 / 27022332
Verifying GPU kernels using range 2276300000000000 to 2276300002000000...
GPU prime generator kernel test passed. (56257 PRPs)
GPU Wieferich/WallSunSun kernel tests passed. (241694743280265 checksum, 0 results)
Starting search...
Workunit complete. Average speed: 23.58 Million primes per second.
19:43:29 (18697): called boinc_finish(0)
shmget in attach_shmem: Invalid argument
19:43:53 (18725): Can't set up shared mem: -1. Will run in standalone mode.
WWocl version 1.03 by Bryan Little and Yves Gallot
Compiled Dec 29 2020 with GCC Apple LLVM 12.0.0 (clang-1200.0.32.28)
A Wieferich and WallSunSun prime number search program
Based on wwwwcl by Mark Rodenkirch and primesieve by Kim Walisch
Command line: ./WW_opencl_macintel64 338772621946054253 338772681946054253
Starting search at: 338772621946054253
Stopping search at: 338772681946054253
GPU Info:
Name: AMD Radeon Pro 5500M Compute Engine
Vendor: AMD
Driver: 1.2 (Oct 29 2020 23:01:05)
Compute Units: 24
Kernel profile: 12.541 ms. Estimated / Actual worksize: 26400000 / 23155161
Verifying GPU kernels using range 338772621946054253 to 338772621948054253...
GPU prime generator kernel test passed. (49688 PRPs)
GPU Wieferich/WallSunSun kernel tests passed. (213679590312100 checksum, 1 results)
Starting search...
Workunit complete. Average speed: 19.22 Million primes per second.
19:45:13 (18725): called boinc_finish(0)
|
5)
Message boards :
Storm message board
(Message 147110)
Posted 889 days ago by McDaWisel
Nice. 10bn! Half of that seems to be from you :-) looking forward to get to 100m myself. Hopefully by the end of January.
Top 10 team finish for 2021 I think!
|
6)
Message boards :
Storm message board
(Message 145465)
Posted 927 days ago by McDaWisel
I am all ramped up!
But I do have one Google cloud instance that's slowly chugging away at 8 x 1 CPU concurrent Cullen tasks :-) They will report in for the first time in about 27 hours.
I think I'll get away with that little boost twice, then I'll have to go to more CPUs per tasks as not to miss the deadline...
|
7)
Message boards :
Storm message board
(Message 145361)
Posted 929 days ago by McDaWisel
Good stuff. We should see the first round of TSC machines reporting in around 12 noon today. That will be interesting :-)
I was aiming for 25 TSC instances, but ended up with 30 as I got more Gold than expected.
Also using up Azure and Google Cloud free credits. Probably the equivalent of another 7 instances.
|
8)
Message boards :
Storm message board
(Message 145338)
Posted 929 days ago by McDaWisel
How seriously is everyone talking this latest challenge? :-)
|
9)
Message boards :
Storm message board
(Message 145047)
Posted 941 days ago by McDaWisel
Thank you! Rather happy with that!
I think I will switch over to the PPSE primes again.... Maybe I'll get lucky there at some stage and get myself on the 5000 list :-)
|
10)
Message boards :
Storm message board
(Message 145044)
Posted 941 days ago by McDaWisel
I think you were right on the money: One had FFT as 2M, the other had FFT as 1920K.
The next two tasks that started are behaving in the same way, so the server must be sending out a mix of these at the moment...
Using zero-padded FMA3 FFT length 2M, Pass1=2K, Pass2=1K, clm=1, 6 threads, a = 3, L2 = 395*370, M = 146150
Using zero-padded FMA3 FFT length 1920K, Pass1=384, Pass2=5K, clm=2, 6 threads, a = 5, L2 = 437*332, M = 145084
|
