sr2sieve client

[mgpower0]

Quick question- Is the sr2sieve client ok to sieve with or does it miss factors?
At the moment I am still running with jjsieve but am also running the same range on the same box with sr2 just as a test. 1.66 c2d runs jj sieve around 324kp/s sr2 running same range at 436kp/s. Much quicker but does it do as good a job is my question

[Joe O]

Originally Posted by mgpower0

Quick question- Is the sr2sieve client ok to sieve with or does it miss factors?
At the moment I am still running with jjsieve but am also running the same range on the same box with sr2 just as a test. 1.66 c2d runs jj sieve around 324kp/s sr2 running same range at 436kp/s. Much quicker but does it do as good a job is my question

SR2sieve uses CPU time to report its speed.
JJsieve uses Wall (elapsed) time to report its speed.

The only way that I know to compare the two, is too take note of the time before and after each program processes a 1G interval when they are running on the same system with all the same things going on, or not going on, as the case may be.

[mgpower0]

yeah im running both on a core2duo jj on 1 core sr2 on the other (both the same range) jjsieve range had about 4hrs start so I guess if sr2 catches up then that will answer part one of the question. Part 2 will sr2 find all the factors. The computer they are running on isn't used for anything apart from DC so noting much else should cut into eithers cpu share

[Citrix]

I don't think sr2sieve misses any factors. As for the speed, I think sr2sieve is faster but you should check it on your machine.

[mgpower0]

yep sr2 overtook jjsieve during the night, and found the same factors so im happy

[engracio]

Can this client also be use in psp or sob sieving? Is it only for linux or ppc? Not really sure. Does is it have windoze client?

[Citrix]

http://www.geocities.com/g_w_reynolds/sr2sieve/

[engracio]

Thanks for the link Citrix. I am running this range on a X2 4600@2640. With jjsieve it was getting 523 kp/s now I guess it is getting 539 kp/s. Not bad of an increase if it does not miss many factors. More importanly am I doing this correctly. Do I just wait for the results and submit it? Thanks.

e


07/07/2007 01:52 PM 73 checkpoint.txt
07/01/2007 09:46 PM 4,498 README.txt
05/06/2007 07:06 AM 13,538,210 SoB.dat
07/06/2007 06:13 AM 86,016 sr2sieve-amd.exe
07/06/2007 06:13 AM 87,040 sr2sieve-intel.exe
07/07/2007 01:52 PM 801 sr2sieve.log
07/07/2007 01:40 PM 15 sr2work.txt
7 File(s) 13,716,653 bytes
2 Dir(s) 94,618,456,064 bytes free
D:\Psr2sieve>sr2sieve-amd -s
sr2sieve 1.5.13 -- A sieve for multiple sequences k*b^n+/-1.
Read 2320262 terms for 18 sequences from dat format file `SoB.dat'.
Split 18 base 2 sequences into 178 base 2^360 subsequences.
Resuming from checkpoint pmin=633500302335713 in `checkpoint.txt'.
Expecting to find factors for about 53.71 terms in this range.
sr2sieve started: 991 <= n <= 49999997, 633500302335713 <= p <= 634000000000000
p=633500723208043, 538998 p/sec, 0 factors, 0.14% done, ETA 18 Jul 08:07

[Citrix]

Originally Posted by engracio

Thanks for the link Citrix. I am running this range on a X2 4600@2640. With jjsieve it was getting 523 kp/s now I guess it is getting 539 kp/s. Not bad of an increase if it does not miss many factors. More importanly am I doing this correctly. Do I just wait for the results and submit it? Thanks.

e


07/07/2007 01:52 PM 73 checkpoint.txt
07/01/2007 09:46 PM 4,498 README.txt
05/06/2007 07:06 AM 13,538,210 SoB.dat
07/06/2007 06:13 AM 86,016 sr2sieve-amd.exe
07/06/2007 06:13 AM 87,040 sr2sieve-intel.exe
07/07/2007 01:52 PM 801 sr2sieve.log
07/07/2007 01:40 PM 15 sr2work.txt
7 File(s) 13,716,653 bytes
2 Dir(s) 94,618,456,064 bytes free
D:\Psr2sieve>sr2sieve-amd -s
sr2sieve 1.5.13 -- A sieve for multiple sequences k*b^n+/-1.
Read 2320262 terms for 18 sequences from dat format file `SoB.dat'.
Split 18 base 2 sequences into 178 base 2^360 subsequences.
Resuming from checkpoint pmin=633500302335713 in `checkpoint.txt'.
Expecting to find factors for about 53.71 terms in this range.
sr2sieve started: 991 <= n <= 49999997, 633500302335713 <= p <= 634000000000000
p=633500723208043, 538998 p/sec, 0 factors, 0.14% done, ETA 18 Jul 08:07

You can look in this thread, the latest version might not be the fastest.
http://www.mersenneforum.org/showthread.php?p=109781

Also, as already mentioned the units for speed measurement in the 2 programs do not correlate. So look for time for a range of billion numbers.

[mgpower0]

i have found sr2 sieve to be quicker on my intels by quite a bit, but on the amd there seems to be little if any difference over jjsieve

[engracio]

Well I took a 1G slice and started both sr2sieve and jjsieve on my e6300 c2d. Let it run all night and see how they compare in the morning.

e