P-1 testing

bayanne · 2022-07-25, 12:07

Is it possible to test p-1 factoring, but using low M_N numbers at large B1 and B2 settings. Also what sort of memory should be allocated.
I am using Mac command line version 30.7b9 so presume that rules out GpuOwl

Prime95 · 2022-07-25, 14:30

It is possible but you must use 30.8b15. Give it all the RAM you can spare as long as the system does not start thrashing during stage 2.

See https://www.mersenneforum.org/showthread.php?t=27477 for exponents to work on.

bayanne · 2022-07-26, 04:59

Quote:

Originally Posted by Prime95

It is possible but you must use 30.8b15. Give it all the RAM you can spare as long as the system does not start thrashing during stage 2.

See https://www.mersenneforum.org/showthread.php?t=27477 for exponents to work on.

But can I use that version with a Mac?

Prime95 · 2022-07-26, 06:30

You could if it existed. I'll see if I can build it this week.

Zhangrc · 2022-07-26, 07:12

Could the PrimeNet server assign separate P-1 work only to version 30.8 or later, and with enough RAM allocated, say 4GB per worker?
Only separate P-1 work at the wavefront should be affected. Those P-1 as a pre-PRP should remain the same.

bayanne · 2022-07-26, 08:46

Quote:

Originally Posted by Prime95

You could if it existed. I'll see if I can build it this week.

Thanks

I can spare 24gb at night and 12gb during day.
Based on this what values of B1 and B2 should I use for an exponent such as M1619 or is that insufficient memory?
Or should I use M5471 or even highed than this?

I intend just running one exponent

Prime95 · 2022-07-26, 13:56

You are highly unlikely to find a factor using P-1 on exponents below 10000. GMP-ECM is also very fast on these tiny numbers and has been available for many years.

I've been working on exponents between 80000 and 100000 using B1=1e9. I've been finding 2 or 3 factors for each exponent range of 1000.

As to memory, I'd let program use 12GB day and night.

bayanne · 2022-07-27, 04:59

Quote:

Originally Posted by Prime95

You are highly unlikely to find a factor using P-1 on exponents below 10000. GMP-ECM is also very fast on these tiny numbers and has been available for many years.

I've been working on exponents between 80000 and 100000 using B1=1e9. I've been finding 2 or 3 factors for each exponent range of 1000.

As to memory, I'd let program use 12GB day and night.

I have no great desire to find primes, so would be happy running on exponents were factors had been found. I just chose those nmubers as examples more than anything else.

Prime95 · 2022-07-28, 01:47

Please try this version: https://www.dropbox.com/s/q4mfqchd4i...SX.tar.gz?dl=0

bayanne · 2022-07-28, 05:09

Quote:

Originally Posted by Prime95

Please try this version: https://www.dropbox.com/s/q4mfqchd4i...SX.tar.gz?dl=0

Thanks

I grabbed half a dozen in the 115m range and now running but in silent mode. Can I view what is being processed?
Should I communicate with you via pm?

bayanne · 2022-07-28, 07:51

Quote:

Originally Posted by bayanne

Thanks

I grabbed half a dozen in the 115m range and now running but in silent mode. Can I view what is being processed?
Should I communicate with you via pm?

CPU temp is running at 100 degrees c so I may decide that that is too warm to run the application ...

How can I show you if the app is running how it should?
Can I run a smaller test?

2022-07-25, 12:07	#1
bayanne "Tony Gott" Aug 2002 Yell, Shetland, UK 2⁴×23 Posts	P-1 testing Is it possible to test p-1 factoring, but using low M_N numbers at large B1 and B2 settings. Also what sort of memory should be allocated. I am using Mac command line version 30.7b9 so presume that rules out GpuOwl

2022-07-25, 14:30	#2
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 1FDE₁₆ Posts	It is possible but you must use 30.8b15. Give it all the RAM you can spare as long as the system does not start thrashing during stage 2. See https://www.mersenneforum.org/showthread.php?t=27477 for exponents to work on. Last fiddled with by Prime95 on 2022-07-25 at 14:32

2022-07-26, 07:12	#5
Zhangrc "University student" May 2021 Beijing, China 414₈ Posts	Could the PrimeNet server assign separate P-1 work only to version 30.8 or later, and with enough RAM allocated, say 4GB per worker? Only separate P-1 work at the wavefront should be affected. Those P-1 as a pre-PRP should remain the same. Last fiddled with by Zhangrc on 2022-07-26 at 07:14

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Status of Wagstaff testing? and testing Mersenne primes for Wagstaff-ness	GP2	Wagstaff PRP Search	414	2020-12-27 08:11
Anti-poverty drug testing vs "high" tax deduction testing	kladner	Soap Box	3	2016-10-14 18:43
k=243 testing ??	gd_barnes	Riesel Prime Search	20	2007-11-08 21:13
Testing	grobie	Marin's Mersenne-aries	1	2006-05-15 12:26
Speed of P-1 testing vs. Trial Factoring testing	eepiccolo	Math	6	2006-03-28 20:53

2022-07-26, 06:30	#4
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2·4,079 Posts	You could if it existed. I'll see if I can build it this week.

2022-07-26, 13:56	#7
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2×4,079 Posts	You are highly unlikely to find a factor using P-1 on exponents below 10000. GMP-ECM is also very fast on these tiny numbers and has been available for many years. I've been working on exponents between 80000 and 100000 using B1=1e9. I've been finding 2 or 3 factors for each exponent range of 1000. As to memory, I'd let program use 12GB day and night.

2022-07-28, 01:47	#9
Prime95 P90 years forever! Aug 2002 Yeehaw, FL 2·4,079 Posts	Please try this version: https://www.dropbox.com/s/q4mfqchd4i...SX.tar.gz?dl=0