PDA

View Full Version : New Prime at PSP



Matt
07-28-2008, 08:47 AM
PSP has found a prime, this means our combined sieve dat now contains 15 Ks. You can get the new dat here: http://www.psp-project.de/test/sievecomb.zip

Zuzu
07-29-2008, 11:27 AM
PSP has found a prime, this means our combined sieve dat now contains 15 Ks. You can get the new dat here: http://www.psp-project.de/test/sievecomb.zip

Congrats to Sloth and to all the PSP community :cheers:. 258317*2^5450519+1 is expected to be the 12th largest prime thusfar. See the thread at www.mersenneforum.org/showthread.php?t=10505

enderak
07-29-2008, 11:55 AM
Wow, Sloth is on a roll over there - especially considering the size of the primes and that it's only been a few months since his last discovery. Congrats!

vjs
07-29-2008, 04:30 PM
Yup I think it was around 120 days between.
For the size of the tests now it's pretty good progress.

Sieve has helped out quite a bit, PSP a little more but also SoB. I wish we could have sieved this deep earlier in the project we would probably be closer to 20M with at least one less k.

In any case since those dats are being sieved together SoB is also close to a 7p sieve depth which is amazing.

Now it's our turn to hand over another prime hopefully in the low 16M.

I personally think we are due another one soon.

vjs
07-29-2008, 04:41 PM
Also it's important to note that they are also doing a pretty good job with a one k approach to secondpass testing.

The idea is to run one k upto or close to the secondpass level while looking for errors and error patterns in the data.

The project co-ordinators then repopulate the secondpass que with these tests. So far their error rate looks good but they are also shorter tests.

SoB should consider doing the same, personally I'd run 22699 upto about 14m and see what happens. If there are obvious problems run those first, if it looks sketchy run 67607 and get a better idea.

Death
07-30-2008, 05:16 AM
yep. we definitely need a priiiiiime.

[DPC]Frentik
08-21-2008, 09:55 AM
Can I update the DAT while sieving or do I have to wait until the current range is finished?

Matt
08-21-2008, 05:34 PM
Probably best to wait until the end of the range just to be safe.

hhh
08-22-2008, 06:36 AM
Probably best to wait until the end of the range just to be safe.

If you stop the client, replace the dat file, and restart the (manual, I suppose) client, no problem. With Boinc, I would let Boinc do the job. H.

[DPC]Frentik
08-25-2008, 05:34 PM
OK thx for the answers. I do use the manual client so I'll update the .dat tomorrow for the remaining 20% of the ranges.