Author |
Message |
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
At the current time, we're planning on restarting the ESP sieve on Saturday after the SR5 challenge ends.
The old PSP sieve (which was actually the PSP/SoB sieve) has been renamed the Sierpinski Problem ESP, PSP, & SoB Sieve. Although the name has changed, it retains the same statistics and badge as before.
From a badge perspective, this is your chance to get or improve one of the "closed" badges.
In terms of life expectancy, I suspect this sieve will run for several years at a minimum.
____________
My lucky number is 75898524288+1 |
|
|
|
Will the PSP/ESP Sieve be enabled by default or will we have to go into our account settings to start running it? |
|
|
|
Great news for those of us that only had slow computers back in the day!
____________
|
|
|
tng Send message
Joined: 29 Aug 10 Posts: 402 ID: 66603 Credit: 23,579,921,500 RAC: 46,905,902
                                      
|
Great news for those of us that only had slow computers back in the day!
Or didn't realize the subproject was about to be suspended -- my one task has been so lonely.
____________
|
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
Will the PSP/ESP Sieve be enabled by default or will we have to go into our account settings to start running it?
It will NOT be enabled by default.
If I remember, I might clear everyone's settings before turning it on. If I forget or decide not to force that setting to "off", if you had PSP Sieve enabled previously it will still be enabled when ESP turns on. I believe that would only happen, however, if you have not changed your settings since PSP was suspended, which was several years ago.
____________
My lucky number is 75898524288+1 |
|
|
|
Should we expect the same credit rate as in the TRP Sieve? |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
Should we expect the same credit rate as in the TRP Sieve?
If I did the math correctly, yes.
____________
My lucky number is 75898524288+1 |
|
|
|
Does PSP/ESP Sieve benefit from enabling muti-threading, like TRP Sieve does?
Edit: and to the same extent? |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
Does PSP/ESP Sieve benefit from enabling muti-threading, like TRP Sieve does?
I have not tried it personally, but I would think the answer is yes.
EDIT: It would probably be to the same extent, yes.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
|
|
Hello !
Will this subproject be CPU only ? or GPU / CPU like PPS Sieve ?
Thank You
Best Regards
Phil1966 |
|
|
Tyler Project administrator Volunteer tester Send message
Joined: 4 Dec 12 Posts: 1077 ID: 183129 Credit: 1,288,886,509 RAC: 406,651
                      
|
Hello !
Will this subproject be CPU only ? or GPU / CPU like PPS Sieve ?
Thank You
Best Regards
Phil1966
This question has been answered before, and the answer is that it is CPU only, sorry no GPU.
____________
275*2^3585539+1 is prime!!! (1079358 digits)
Proud member of Aggie the Pew
|
|
|
|
At the current time, we're planning on restarting the ESP sieve on Saturday after the SR5 challenge ends.
Is there any change of plans or can we still expect the new sieve today?
____________
676754^262144+1 is prime |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
At the current time, we're planning on restarting the ESP sieve on Saturday after the SR5 challenge ends.
Is there any change of plans or can we still expect the new sieve today?
I'm not sure yet. I spent most of today dealing with a connectivity problem between my home network and Rackspace. I haven't had a chance to look at the sieve yet today.
So the answer is... maybe.
I'll definitely have an answer for you, one way or another, later today.
____________
My lucky number is 75898524288+1 |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
At the current time, we're planning on restarting the ESP sieve on Saturday after the SR5 challenge ends.
Is there any change of plans or can we still expect the new sieve today?
I'm not sure yet. I spent most of today dealing with a connectivity problem between my home network and Rackspace. I haven't had a chance to look at the sieve yet today.
So the answer is... maybe.
I'll definitely have an answer for you, one way or another, later today.
It looks like we are going to turn it on later tonight. That's tonight, my time, as it's already Sunday UTC. Probably about 2 to 3 hours from now if nothing goes sideways.
____________
My lucky number is 75898524288+1 |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
ESP Sieve is now live.
Go ahead and run it. Let us know if there's any problems.
____________
My lucky number is 75898524288+1 |
|
|
RogerVolunteer developer Volunteer tester
 Send message
Joined: 27 Nov 11 Posts: 1137 ID: 120786 Credit: 267,656,892 RAC: 7,886
                    
|
Stats look like they are reporting correctly:
http://stats.free-dc.org/stats.php?page=userwork&proj=pgrid&subproj=PSP_sr2sieve&sort=today
Looks like I am the third person to have a WU validate. |
|
|
|
I've got 2 tasks finally validated and 34 pending. It seems that the sieve yields a little bit more credits than the TRP Sieve.
Also the sieve seems to be very prolific as my two accomplished tasks resulted in 74 factors. Is there any estimation on the optimal sieve depth?
Thanks to Michael and the team for restarting the sieve and giving us a chance to earn/upgrade the badge. |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
I've got 2 tasks finally validated and 34 pending. It seems that the sieve yields a little bit more credits than the TRP Sieve.
Slightly more credit per task, but should be almost the exact same credits/hour as the TRP sieve.
Also the sieve seems to be very prolific as my two accomplished tasks resulted in 74 factors.
That's due to the current depth of the sieve. The deeper you are, the fewer factors you find.
The ESP sieve started at 5T (0.005P). The TRP sieve, by comparison, is currently at 35867T (35.867P). Much lower P, much higher density of factors.
Don't get confused by the two different P's there. The sieve depth is the variable called P. However, P, when used in a number, means "Peta", which is numerical shorthand for 10^15, which is the same as adding 15 zeros to the number. So if I write P=35P, that means that the sieve depth is 35000000000000000.
Also, the ESP sieve yields more factors per task because the ESP sieve tasks are sieving a larger range. The ESP sieve is faster than the TRP sieve (because there's a lot fewer k's to search.) We have therefore set each ESP task to sieve a larger P range than the TRP sieve tasks. TRP sieves 2G per task while ESP sieves 5G per task.
Is there any estimation on the optimal sieve depth?
I did some rough calculation in my head, which may be wrong. Since we're sieving for a rather large range of n (10M to 50M), the optimal sieve depth is different for the start of the range (10M) than it is for the end of the range (50M). It will be a VERY long time before we LLR the numbers at the end of the range. SoB, which is at n=27M, is barely halfway there.
At the beginning of the range (n=10M), the optimal sieving point is between 5P and 10P, and my estimate is it will take between 200 and 400 days to get there.
The optimal sieving point for the end of the range (n=50M) is between 163P and 327P, and will take between 6000 and 12000 days to get to the optimal depth.
Obviously we won't be stopping this sieve any time soon.
____________
My lucky number is 75898524288+1 |
|
|
|
Hello Michael,
Thanks to Michael and the team for restarting the sieve and giving us a chance to earn/upgrade the badge.
+1 !
It works very well !!!
Thank you !!!
9 WU's validated
96 Waiting validation
Best Regards
Phil1966 |
|
|
quelSend message
Joined: 21 Nov 10 Posts: 86 ID: 73895 Credit: 455,877,161 RAC: 51,844
                   
|
Everything looks good from here as well. So far 4 machines with WUs that have validated which includes a mix of win and lin across i7s and 1 old phenom II X6. No errors and no invalids.
I am looking forward to getting a crack at badges for this one. |
|
|
|
Thank you Michael for the response.
Another question is why don't we sieve in smaller ranges, say 10M-15M and stop at optimal sieve depth? It means we won't be "overseiving" lower ranges. |
|
|
JimB Honorary cruncher Send message
Joined: 4 Aug 11 Posts: 916 ID: 107307 Credit: 974,494,172 RAC: 911
                    
|
Another question is why don't we sieve in smaller ranges, say 10M-15M and stop at optimal sieve depth? It means we won't be "overseiving" lower ranges.
It's more efficient to sieve a larger n range at once. The n=50M end point was chosen because PSP and SoB are already sieved to that point. At some future time we could add PSP and SoB candidates back to the sieve, if needed.
The sieve file can always be changed to increase the minimum n (either because it's optimally sieved or because LLR testing has reached that high), to remove a k because a prime was found or just to remove all the candidates where factors were found. I see we've changed the TRP sieve file eleven times since that project started. This sieve file will most definitely undergo periodic changes. |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 2838 ID: 130544 Credit: 974,577,702 RAC: 1,128,993
                      
|
Our little Atom N270 just resolved this _13 quora unit:
http://www.primegrid.com/result.php?resultid=562383920 |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 2838 ID: 130544 Credit: 974,577,702 RAC: 1,128,993
                      
|
Every PSP sv I now get has a highish quora value - hardly getting any fresh.
Not complaining because it's sieve + unlike LLR it's not like one can be 'only be the double-checker' :).
Has a big-hitter's cache blown up recently or something? |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
Every PSP sv I now get has a highish quora value - hardly getting any fresh.
Not complaining because it's sieve + unlike LLR it's not like one can be 'only be the double-checker' :).
Has a big-hitter's cache blown up recently or something?
I don't know. Pure speculation: Perhaps a lot of tasks got aborted as people switched to TRP-Sieve for the challenge? The other alternative would be that there's something wrong with the validator, although nothing has changed lately so that would be unexpected. Please let me know if *your* PSP-Sieve tasks are validating and getting credit.
____________
My lucky number is 75898524288+1 |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
I looked at a few of those workunits. From the perspective of the validation system, nothing looks wrong. There's just a lot of aborted tasks -- and most of them seem to be on the day the challenge started.
____________
My lucky number is 75898524288+1 |
|
|
tng Send message
Joined: 29 Aug 10 Posts: 402 ID: 66603 Credit: 23,579,921,500 RAC: 46,905,902
                                      
|
I looked at a few of those workunits. From the perspective of the validation system, nothing looks wrong. There's just a lot of aborted tasks -- and most of them seem to be on the day the challenge started.
There was a large increase in the amount of work being done, apparently as a result of the challenge mentioned in this thread. That challenge would have ended at about that time.
____________
|
|
|
|
This "The Sierpinski Problem ESP/PSP/SoB Sieve (ESP-Sieve)" project is running on one of my computers without being selected in my Primegrid preferences. Even if I deselect it in my "School" profile, it comes back.
Is there a way to solve this problem ? |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
This "The Sierpinski Problem ESP/PSP/SoB Sieve (ESP-Sieve)" project is running on one of my computers without being selected in my Primegrid preferences. Even if I deselect it in my "School" profile, it comes back.
Is there a way to solve this problem ?
At the top of the page where you select the tasks there are three boxes, "Use CPU", "Use ATI GPU", and "Use Nvidia GPU". You have the "Use CPU" and "Use ATI GPU" boxes checked.
When the "Use CPU" box is checked, the server will send CPU tasks to your computer. If you don't select which CPU tasks you want to get, the server selects a task for you -- in this case, it selects the ESP Sieve.
As long as that "Use CPU" box is checked, you'll keep getting CPU tasks.
If you don't want to get any CPU tasks, uncheck the "Use CPU" box.
____________
My lucky number is 75898524288+1 |
|
|
|
I cannot seem to be able to this project off |
|
|
|
I cannot seem to be able to this project off
Hi James Lee, I'm not sure I understand your question but I'll mention a few things that might help. There are two "extended sierpinski problem" subprojects here, one "LLR" (for direct prime-finding) and one "sieve" (to reduce the number of LLR candidates). Both run on CPU only; GPU is not useful for these tasks. Look at your Preferences page (click the "preferences" button in the BOINC GUI manager) and it will open in your web browser. To run one or the other ensure the "use CPU" box is checked, then scroll down to select the subprojects that you want to run, then save the settings.
If you're trying to disable PrimeGrid, simply do so from the "Projects" tab of the BOINC manager window. Obviously you have had some success running PrimeGrid, since you have credit in the GFN subproject. But, your computers are hidden which makes remote debugging a bit difficult. Good luck and post back here (or PM me) if you need help. Cheers.
--Gary |
|
|
JimB Honorary cruncher Send message
Joined: 4 Aug 11 Posts: 916 ID: 107307 Credit: 974,494,172 RAC: 911
                    
|
I cannot seem to be able to this project off
If you don't select a CPU project, ESP sieve will be selected for you. If you don't want any CPU projects, uncheck Use CPU near the top. |
|
|
Michael Goetz Volunteer moderator Project administrator
 Send message
Joined: 21 Jan 10 Posts: 13520 ID: 53948 Credit: 242,417,032 RAC: 338,482
                          
|
In terms of life expectancy, I suspect this sieve will run for several years at a minimum.
We restarted this sieve about 18 months ago, thinking that we'd need a new ESP sieve file in about 2 years' time. We were hoping we'd be able to get enough new sieving done in those two years to bring ESP up to the same level asp PSP/SoB, or at least be fairly close.
It turns out that two years was a decent estimate for both LLR and the sieve.
LLR is currently just shy of n=9M. The old sieve file goes up to n=10M. When we moved ESP to BOINC it was at about n=5M.
The sieving target is 90P (same as PSP//SoB), and we're currently at 69P.
Although no decision has been made yet, the possibility exists that the ESP sieve will be suspended when it gets to 90P. That should happen in about 6 to 12 months, depending on participation. If you care about badges, consider yourself warned that this sub-project may shut down later this year.
If it does shut down, we will give you at least 1 month's advance notice.
____________
My lucky number is 75898524288+1 |
|
|
|
Thanks for the heads-up. |
|
|
Dave  Send message
Joined: 13 Feb 12 Posts: 2838 ID: 130544 Credit: 974,577,702 RAC: 1,128,993
                      
|
Wowsers. Working towards turq as we speak. |
|
|
|
Although no decision has been made yet, the possibility exists that the ESP sieve will be suspended when it gets to 90P. That should happen in about 6 to 12 months, depending on participation. If you care about badges, consider yourself warned that this sub-project may shut down later this year.
It will be interesting to see how that works out with the Summer Paralympics Challenge on ESP-Sieve essentially seven months away. It may turn out to be a going away challenge for the project until we need to revive it to push the sieve deeper (or unless we decide to push the sieve deeper immediately rather than retire it).
____________
|
|
|