GPU Trial Factoring FAQ - Page 3

Bdot · 2011-11-07, 23:02

Dubslow mentioned I might want to comment on the mfakto-part of the pdf ... sorry I did not take part in that earlier.

And indeed, the "DO NOT SUBMIT “NO FACTOR” RESULTS YET" statement is a bit outdated. Apart from the bug with version 0.08 that may lead to missed factors < 2⁴⁸, there's no known reason for not reporting all results.

Regarding the Factor size: There is no lower limit for the factors, so for mfaktc this should just read "Factor size <= 2⁹⁵ ". mfakto does not yet have the 95-bit kernel, so for mfakto the limit is "Factor size <= 2⁹² ".

As an estimated GIMPS score you can use (maybe I should test that some day):

roughly 60 GHz days/day on HD5770 & 2 CPU cores

This is also a bit misleading: "mfakto works best on large exponents as the sieving (CPU) part stays constant with larger numbers."
mfakto (and mfaktc) are more efficient with larger assignments. This does not mean larger exponents, but "more work" or "longer runtime per class", generally bigger bit ranges. The reason is a certain one-time initialization effort per class - no matter if the class will just test 1 million factor candidates or 1 billion. In the first case, the one-time effort may account for, say, 75% of the sieving effort, in the latter case just 0.3%. Plus there is an average of half a block wasted for each class. If the class consists of only one block, that's 50%, for 1000 blocks it's just 0.05%.

Brain · 2011-11-08, 06:10

I'm going to integrate this information asap.

Uncwilly · 2011-11-08, 07:14

Quote:

Originally Posted by Brain

I'm going to integrate this information asap.

You may want to add about the mini-pirmenet that Chasall has set up to aid the factoring to extra bit depths ahead of the LL 'wave front'.

Bdot · 2011-11-08, 08:39

Quote:

Originally Posted by Bdot

As an estimated GIMPS score you can use (maybe I should test that some day):

roughly 60 GHz days/day on HD5770 & 2 CPU cores

I just checked the crunching-power of my HD5770-box. The last 24 hrs returned 96.6 GHz-days, using 3 CPU-cores. So I guess mentioning 90 GHz-days/day with 2 cores should be quite safe.

Brain · 2011-11-08, 21:26

This is a replacement for 0.05a. Please review.

Several minor changes:

mfakto: All results submittable
mfakto: Decreased upper factor size limit
GPU to 72 tool
ATI driver warning for 11.10
ATI: GIMPS score estimate added
CUDALucas: Unknown upper limit
Call for P-1 and ECM implementations

GIMPS GPU Computing Cheat Sheet (pdf)

LaurV · 2011-11-09, 04:31

Quote:

Originally Posted by Brain

P.S.: We still need somebody to implement the following work types on a GPU:
-- P-1 factoring
-- ECM factoring
Code it and enter the GIMPS hall of fame…

How about CudaNFS, CudaQS/SIQS/MPQS (at least for init phase)?

Will it have some effect on the speed(time)? etc.

xilman · 2011-11-12, 11:42

Quote:

Originally Posted by LaurV

How about CudaNFS, CudaQS/SIQS/MPQS (at least for init phase)?

Will it have some effect on the speed(time)? etc.

Undoubtedly possible, but doubtedly worth the effort for NFS. Assuming, that is, you want the sieving phase to be implemented in CUDA. Polynomial selection is already available in msieve. Filtering is too I/O intensive to make sense to port to CUDA and I don't yet have a good enough feel for the square root phase to make sensible comments.

As a rule, GPUs are good at computation, poor at memory access and lousy at I/O compared with CPUs. The sieving phase is very heavily memory bound ...

Paul

LaurV · 2011-11-13, 06:17

Thanks for the answer, it was more like a curiosity, my understanding of NFS is almost zero at this time.

Dubslow · 2011-11-13, 06:26

Quote:

Originally Posted by Brain

This is a replacement for 0.05a. Please review.

Several minor changes:

mfakto: All results submittable
mfakto: Decreased upper factor size limit
GPU to 72 tool
ATI driver warning for 11.10
ATI: GIMPS score estimate added
CUDALucas: Unknown upper limit
Call for P-1 and ECM implementations

garo...
or some other mod...
(and the other FAQ thread as well)

Dubslow · 2012-01-11, 22:05

Looking for an old post, found that this PDF is still v.06, while the other FAQ has v.07.

Bdot · 2012-01-13, 09:27

BTW, mfakto 0.10 / 0.10p1 solved the issues with the latest Catalyst drivers. 11.10 to 11.12 are tested well.

2011-11-07, 23:02	#23
Bdot Nov 2010 Germany 3×199 Posts	mfakto update, please Dubslow mentioned I might want to comment on the mfakto-part of the pdf ... sorry I did not take part in that earlier. And indeed, the "DO NOT SUBMIT “NO FACTOR” RESULTS YET" statement is a bit outdated. Apart from the bug with version 0.08 that may lead to missed factors < 2⁴⁸, there's no known reason for not reporting all results. Regarding the Factor size: There is no lower limit for the factors, so for mfaktc this should just read "Factor size <= 2⁹⁵ ". mfakto does not yet have the 95-bit kernel, so for mfakto the limit is "Factor size <= 2⁹² ". As an estimated GIMPS score you can use (maybe I should test that some day): roughly 60 GHz days/day on HD5770 & 2 CPU cores This is also a bit misleading: "mfakto works best on large exponents as the sieving (CPU) part stays constant with larger numbers." mfakto (and mfaktc) are more efficient with larger assignments. This does not mean larger exponents, but "more work" or "longer runtime per class", generally bigger bit ranges. The reason is a certain one-time initialization effort per class - no matter if the class will just test 1 million factor candidates or 1 billion. In the first case, the one-time effort may account for, say, 75% of the sieving effort, in the latter case just 0.3%. Plus there is an average of half a block wasted for each class. If the class consists of only one block, that's 50%, for 1000 blocks it's just 0.05%. Last fiddled with by Bdot on 2011-11-07 at 23:41 Reason: large exponents vs. large assignments

2011-11-08, 06:10	#24
Brain Dec 2009 Peine, Germany 331 Posts	Thanks I'm going to integrate this information asap.

2011-11-08, 21:26	#27
Brain Dec 2009 Peine, Germany 331 Posts	GPU Computing Guide Update to v 0.06 This is a replacement for 0.05a. Please review. Several minor changes: mfakto: All results submittable mfakto: Decreased upper factor size limit GPU to 72 tool ATI driver warning for 11.10 ATI: GIMPS score estimate added CUDALucas: Unknown upper limit Call for P-1 and ECM implementations GIMPS GPU Computing Cheat Sheet (pdf) Last fiddled with by Brain on 2012-08-05 at 10:15

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
What is Trial Factoring?	Unregistered	Information & Answers	5	2012-08-02 03:47
How much Trial Factoring to do?	odin	Software	4	2010-08-08 20:23
How far to do trial factoring	S485122	PrimeNet	1	2007-09-06 00:52
How to only do Trial Factoring?	michael	Software	23	2004-01-06 08:54
About trial factoring	gbvalor	Math	4	2003-05-22 02:04

2011-11-13, 06:17	#30
LaurV Romulan Interpreter "name field" Jun 2011 Thailand 41·251 Posts	Thanks for the answer, it was more like a curiosity, my understanding of NFS is almost zero at this time.

2012-01-11, 22:05	#32
Dubslow Basketry That Evening! "Bunslow the Bold" Jun 2011 40<A<43 -89<O<-88 1110000110101₂ Posts	Looking for an old post, found that this PDF is still v.06, while the other FAQ has v.07.

2012-01-13, 09:27	#33
Bdot Nov 2010 Germany 3·199 Posts	BTW, mfakto 0.10 / 0.10p1 solved the issues with the latest Catalyst drivers. 11.10 to 11.12 are tested well.