Help test 2995125705 - Find a new top 5000 prime! - Page 10

segmtfault · 2004-12-21, 17:37

Congratulations about your prime !

I am reserving 496-500

lsoule · 2004-12-22, 00:11

484to488 complete, no primes.

Reserving 500to504.

lsoule · 2004-12-23, 06:38

500-504 complete, no primes.
Reserving 504-508

Kosmaj · 2004-12-23, 09:45

Larry,
I noticed that your result files are interlaced with "Iteration" output, like
Iteration 363607 / 488002

which are not so useful, and they reset the time so each candidate has different exe time recorded. I suspect you have a small value set for "ResultsFileIterations", can you check in your "Preferences" and set that value to a large number, like 99999999 (which is the default).

Then can you tell us is there any "jump" in exe times at about n=505k, I think the FFT length will soon change from 49152 to 57344.

BTW, I updated the table and I added values to 540k.

segmtfault · 2004-12-23, 15:34

Christmas vacation reservation :
I am taking 508-512, 512-516 and 516-520
I'll be back on January 8th. Merry Christmas and happy new year

lsoule · 2004-12-23, 17:32

The iteration time goes up 20% when going from fftlen of
49152 to 57344.

The reason for the iteration line showing up is that I farm the jobs
out onto unloaded machines and run for 10-15 minutes. After that
time, I kill it and switch it to the lowest loaded machine. A lot
of jobs here come and go so sometimes a machine gets loaded quickly...
I can increase the time it runs so that some exponents finish all in one run.

-Larry

ValerieVonck · 2004-12-24, 08:44

I will take520to524.txt

CedricVonck [Reserved]

lsoule · 2004-12-24, 14:59

Reserving 524to528 and 528to532

Kosmaj · 2004-12-24, 15:54

CedricVonck
Welcome to the group! I'm not sure how familiar you are with prime number search and how many have you found so far, I noticed you searched for some "prime Seirpinski" numbers, in any case if you have any questions don't hesitate to ask. Just remember two things, (1) if you find a prime please include in your prover's code "15k" search, and (2) mail me the results file when you are done. You can find the details in the first post of the thread.

BTW, I reconstructed the file with all candidates to n=1M sieved to 3250 bn. There are quite a few candidates (about 155 per 1000) remaining and if we go on with testing I suggest we do some more sieveing, to 4000 or 5000 bn. We can arrange the details early in January. Are there any volunteers? It's the best to sieve using "ksieve" on a fast Athlon or on 1GHz or faster P-3 letting P-4's run LLR.

Merry Christmas!

lsoule · 2004-12-24, 23:45

I'd be up for sieving if I can get my hands on a copy of ksieve that
runs on linux (and that I can access from the US

)

em99010pepe · 2004-12-27, 22:28

I can put my AMD to sieve, just explain me how, step by step.

Carlos

2004-12-27, 22:28	#110
em99010pepe Sep 2004 2·5·283 Posts	I can put my AMD to sieve, just explain me how, step by step. Carlos Last fiddled with by em99010pepe on 2004-12-27 at 22:35

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Choose your own K and work on finding a top-5000 prime!	lsoule	Riesel Prime Search	421	2021-07-27 23:51
Test a Specific Exponent in Prime95. When it is 100% done, does it mean I find a new prime number?	king	Information & Answers	5	2018-02-21 18:15
If you find a prime...	paulunderwood	NeRDs	0	2014-02-03 05:09
Help test 210885 - Find a new top 5000 prime!	SlashDude	Riesel Prime Search	121	2008-01-03 08:47
The fastest way to a top-5000 prime?	lsoule	15k Search	13	2005-09-19 20:24

2004-12-21, 17:37	#100
segmtfault Jun 2004 California 115₁₆ Posts	Congratulations about your prime ! I am reserving 496-500

2004-12-22, 00:11	#101
lsoule Nov 2004 California 2³×3×71 Posts	484to488 complete, no primes. Reserving 500to504.

2004-12-23, 06:38	#102
lsoule Nov 2004 California 2³×3×71 Posts	500-504 complete, no primes. Reserving 504-508

2004-12-23, 09:45	#103
Kosmaj Nov 2003 2×1,811 Posts	Larry, I noticed that your result files are interlaced with "Iteration" output, like Iteration 363607 / 488002 which are not so useful, and they reset the time so each candidate has different exe time recorded. I suspect you have a small value set for "ResultsFileIterations", can you check in your "Preferences" and set that value to a large number, like 99999999 (which is the default). Then can you tell us is there any "jump" in exe times at about n=505k, I think the FFT length will soon change from 49152 to 57344. BTW, I updated the table and I added values to 540k.

2004-12-23, 15:34	#104
segmtfault Jun 2004 California 277 Posts	Christmas vacation reservation : I am taking 508-512, 512-516 and 516-520 I'll be back on January 8th. Merry Christmas and happy new year

2004-12-23, 17:32	#105
lsoule Nov 2004 California 1704₁₀ Posts	The iteration time goes up 20% when going from fftlen of 49152 to 57344. The reason for the iteration line showing up is that I farm the jobs out onto unloaded machines and run for 10-15 minutes. After that time, I kill it and switch it to the lowest loaded machine. A lot of jobs here come and go so sometimes a machine gets loaded quickly... I can increase the time it runs so that some exponents finish all in one run. -Larry

2004-12-24, 08:44	#106
ValerieVonck Mar 2004 Belgium 29² Posts	I will take520to524.txt CedricVonck [Reserved]

2004-12-24, 14:59	#107
lsoule Nov 2004 California 2³×3×71 Posts	Reserving 524to528 and 528to532

2004-12-24, 15:54	#108
Kosmaj Nov 2003 2·1,811 Posts	CedricVonck Welcome to the group! I'm not sure how familiar you are with prime number search and how many have you found so far, I noticed you searched for some "prime Seirpinski" numbers, in any case if you have any questions don't hesitate to ask. Just remember two things, (1) if you find a prime please include in your prover's code "15k" search, and (2) mail me the results file when you are done. You can find the details in the first post of the thread. BTW, I reconstructed the file with all candidates to n=1M sieved to 3250 bn. There are quite a few candidates (about 155 per 1000) remaining and if we go on with testing I suggest we do some more sieveing, to 4000 or 5000 bn. We can arrange the details early in January. Are there any volunteers? It's the best to sieve using "ksieve" on a fast Athlon or on 1GHz or faster P-3 letting P-4's run LLR. Merry Christmas!

2004-12-24, 23:45	#109
lsoule Nov 2004 California 3250₈ Posts	I'd be up for sieving if I can get my hands on a copy of ksieve that runs on linux (and that I can access from the US )