Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
Message boards :
The Riesel Problem :
TRP-Sieve tasks to increase in size
| Author |
Message |
|---|
 Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
                           
|
|
We are going to increase the length of the TRP-Sieve tasks by a factor of 5. Current tasks sieve 2G and the new tasks will sieve 10G. Both the run times and the credit will increase 5x compared to the current tasks.
These tasks used to be 10G in size, but were shortened to accommodate Android devices. Since the Android app has been broken for years, and there's nobody working on fixing it, there's no point in keeping these tasks artificially small anymore.
We anticipate making this change sometime after the SGS challenge completes on Monday.
____________
My lucky number is 75898524288+1 | |
|
RafaelVolunteer tester
Send message
Joined: 22 Oct 14
Posts: 885
ID: 370496
Credit: 334,085,845
RAC: 0
            
|
|
I kinda liked tasks being short... but oh well, I guess 5~6h is at least more manageable than 20~24. | |
|
|
|
|
I thought you were discontinuing the TRP-Sieve task. Have I been crunching the wrong task these last few weeks? | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
I thought you were discontinuing the TRP-Sieve task. Have I been crunching the wrong task these last few weeks?
You are not mistaken, but don't forget that the discussion about TRP-Sieve ending is very tentative and nothing is decided yet. It is still possible that it will end sooner than expected, but nothing is certain, and just how soon "soon" actually might be is unknown. Also, increasing task size (to an even larger 40G) was part of the original discussion about ending the sieve.
In order to keep related information in one place, please ask further questions about the sieve potentially ending in the big "the sieve is ending" thread, while questions about the change in task size should be asked here in this thread. Thank you!
____________
My lucky number is 75898524288+1 | |
|
|
|
|
I do not mind the 10g. size task. most of my climate@home are from 15 to 150 days in length. the task is not so bad as long is used for something that will help the earth.
____________
| |
|
|
|
|
I prefer increasing the size. :-) | |
|
|
|
We are going to increase the length of the TRP-Sieve tasks by a factor of 5. Current tasks sieve 2G and the new tasks will sieve 10G. Both the run times and the credit will increase 5x compared to the current tasks.
These tasks used to be 10G in size, but were shortened to accommodate Android devices. Since the Android app has been broken for years, and there's nobody working on fixing it, there's no point in keeping these tasks artificially small anymore.
We anticipate making this change sometime after the SGS challenge completes on Monday.
Although I personally have almost zero programming experience and therefore can't volunteer, I would gladly connect my Android devices if the program was corrected in the future. I understand the desire to change now absent the android program working. | |
|
|
|
...We anticipate making this change sometime after the SGS challenge completes on Monday.
Monday? Or perhaps Tuesday?!?
| |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
...We anticipate making this change sometime after the SGS challenge completes on Monday.
Monday? Or perhaps Tuesday?!?
Indeed, it's Tuesday.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
There are so many prime number projects. Do any of them overlap the range of the TRP-Sieve? Is there a possibility of say a GFN prime also being a TRP prime, or any other? I remember reading about GFN 15-16 being at the end of their life and more work needs to be done on the 17's which could help other projects. Where is the greatest need in prime grid? Is it in the sieving or actual looking for primes? | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
There are so many prime number projects. Do any of them overlap the range of the TRP-Sieve? Is there a possibility of say a GFN prime also being a TRP prime, or any other? I remember reading about GFN 15-16 being at the end of their life and more work needs to be done on the 17's which could help other projects. Where is the greatest need in prime grid? Is it in the sieving or actual looking for primes?
From time to time, there's something really important and high priority. The recently completed ESP, for example. We needed to do that sieving to support ESP-LLR which had moved from PRPNet to BOINC and was seeing much higher volume. In the near future, we'll also move GCW from PRPNet to BOINC and as a result GCW sieving will be a high priority of ours. But at the current time, there's no pressing need to do any particular project very quickly.
tl;dr: The projects do not overlap.
As for projects overlapping, they never search the same numbers. All of the GFN projects, SGS, SR5, Cullen, Woodall, ESP, PSP, and SoB, TRP and GCW will never overlap any other projects.
GFN won't overlap with other projects because b is so large.
SGS, Cullen, Woodall, and GCW won't overlap because k is so large.
GCW and SR5 won't overlap because they have unique b values.
ESP, PSP, and SoB won't overlap because their remaining k values are larger than those in PPSE.
TRP has different k values than any other project that searches numbers with c = -1.
The remaining projects (PPS, PPSE, PPS-MEGA, 321, 27, and 121) are all variations on Proth numbers. (The last three also do both Proth and Riesel numbers, and the following logic also applies on the Riesel side). We organize the candidates for each project such that they don't search the same numbers. The PPS search, for example, excludes k values of 3, 27, and 121, which are searched separately in the 321, 27, and 121 projects.
So, either the numbers are, by definition, so very different that there's no way for the searches to overlap, or in a few cases we've specifically selected the candidate ranges so that they won't overlap.
____________
My lucky number is 75898524288+1 | |
|
JimBHonorary cruncher
Send message
Joined: 4 Aug 11
Posts: 916
ID: 107307
Credit: 974,514,092
RAC: 0
    
|
The PPS search, for example, excludes k values of 3, 27, and 121, which are searched separately in the 321, 27, and 121 projects.
The PPS search does not search k=3, but does search k=27 and k=121. The residues from the PRPNet search are imported into BOINC, so that only one test is started on k=27 and k=121. If that one doesn't match the PRPNet residue, only then is a second BOINC test required.
| |
|
|
|
GFN won't overlap with other projects because b is so large.
And what about overlapping between different GFN projects? For example, is it posible that, in a given time, GFN 15 to overlap GFN 16? | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
GFN won't overlap with other projects because b is so large.
And what about overlapping between different GFN projects? For example, is it posible that, in a given time, GFN 15 to overlap GFN 16?
Theoretically possible for a small number of candidates, but we remove those candidates that appear in a second range.
For example,
4^32768+1 is the same number as 2^65536+1.
____________
My lucky number is 75898524288+1 | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13513
ID: 53948
Credit: 237,712,514
RAC: 0
|
|
The new super-sized TRP sieve tasks are now live. It will take a little while for the new tasks to work their way though the result queue, so I expect they will start getting sent out in about 2 hours.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
My first batch came through six hours ago. | |
|
|
|
|
I have seen some tasks are double the runtime as older tasks but I am seeing the same credit.. should I?
____________
Thanks,
Jim
 | |
|
|
|
|
Never mind. Found the answer.
____________
Thanks,
Jim
| |
|
Message boards :
The Riesel Problem :
TRP-Sieve tasks to increase in size |
