Weird results of Msieve

jacky · 2017-10-09, 03:00

I have tested to factor C100 on two machines with GPU:

USE_CUDA = True

Command:"..\factMsieve.py example.n"

1、CPU:i7 3.60GHz RAM：8G GTX 745 elapsed time 7hours

2、CPU:E5-2670 2.60GHz RAM:64G Tesla P100-PCIE-16GB elapsed time 16hours

The result confuesd me.It seems that GPU card doesn't work.Why Tesla is much slower?

And if only test the first step,the command "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y" -t 4 -s gpu0polyfile"got the same result of "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y"-g 0 -t 8 -s gpu0polyfile".It seems both GPU and multi threads do not work.How to choose the number of thread?(-t number)

wombatman · 2017-10-09, 03:04

Quote:

Originally Posted by jacky

I have tested to factor C100 on two machines with GPU:

USE_CUDA = True

Command:"..\factMsieve.py example.n"

1、CPU:i7 3.60GHz RAM：8G GTX 745 elapsed time 7hours

2、CPU:E5-2670 2.60GHz RAM:64G Tesla P100-PCIE-16GB elapsed time 16hours

The result confuesd me.It seems that GPU card doesn't work.Why Tesla is much slower?

And if only test the first step,the command "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y" -t 4 -s gpu0polyfile"got the same result of "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y"-g 0 -t 8 -s gpu0polyfile".It seems both GPU and multi threads do not work.How to choose the number of thread?(-t number)

Leave off "-nps". Only "-np1" step is GPU-enabled. The "-nps -npr" steps are CPU only and single-threaded only.

jacky · 2017-10-09, 03:21

Quote:

Originally Posted by wombatman

Leave off "-nps". Only "-np1" step is GPU-enabled. The "-nps -npr" steps are CPU only and single-threaded only.

Thanks.
For the first result,why Tesla is slower if using python to factor.Is it the same reason?

wombatman · 2017-10-09, 04:07

Quote:

Originally Posted by jacky

Thanks.
For the first result,why Tesla is slower if using python to factor.Is it the same reason?

You won't be able to tell if the Tesla is actually slower until you run the "-np1" step by itself. If a given CPU is slower, then running the CPU-bound -nps and -npr steps would make the overall process slower.

VBCurtis · 2017-10-09, 04:18

I don't think a GPU is any faster than CPU for a 100-digit input. That job is so small that invoking the GPU at all is a waste of effort; you may not be able to tell which is faster when the GPU works for something like 2 minutes.

GPU-enhanced poly select doesn't become notably more powerful than CPU until 140 digits or so.

jacky · 2017-10-09, 06:57

Quote:

Originally Posted by VBCurtis

I don't think a GPU is any faster than CPU for a 100-digit input. That job is so small that invoking the GPU at all is a waste of effort; you may not be able to tell which is faster when the GPU works for something like 2 minutes.

GPU-enhanced poly select doesn't become notably more powerful than CPU until 140 digits or so.

Thanks.I seem to have found the problem according to your answer.
Now I am testing a 155-digit number for np1 on two machines with the same command"msieve.exe -np1 -g 0 -t 4 -s gpu0polyfile".I want to know if GPU works.
Is it right? How many hours it usually takes for np1?

jacky · 2017-10-09, 07:27

Now Tesla 100 is 4 times faster than GTX 745.It seems not to have achieved the desired result.I think Tesla should be much faster.Should I give more threads?

chris2be8 · 2017-10-09, 16:08

Here is the log from my run to generate a poly for a C167, which took 7:19:19. Note I only ran -npr on the best 200 lines output by -np1 -nps. Limiting the range of leading coefficients to search would speed it up.

I've also attached the perl script I ran to generate it. It's designed to work on Linux and requires UNIX utilities sort and tail. But it might work on a UNIX like environment under Windows. You would need to update the paths to resources in it, particularly GPUSORT and GPUPTX.

I hope you find it useful.

Chris

jacky · 2017-10-10, 01:41

Quote:

Originally Posted by chris2be8

Here is the log from my run to generate a poly for a C167, which took 7:19:19. Note I only ran -npr on the best 200 lines output by -np1 -nps. Limiting the range of leading coefficients to search would speed it up.

I've also attached the perl script I ran to generate it. It's designed to work on Linux and requires UNIX utilities sort and tail. But it might work on a UNIX like environment under Windows. You would need to update the paths to resources in it, particularly GPUSORT and GPUPTX.

I hope you find it useful.

Chris

Thank you very much.It's very useful to me.But I have a question,if limiting the range of leading coefficients to search,will it lead to failure in polynomial selection ?

VBCurtis · 2017-10-10, 05:47

Quote:

Originally Posted by jacky

Thank you very much.It's very useful to me.But I have a question,if limiting the range of leading coefficients to search,will it lead to failure in polynomial selection ?

Nope.

chris2be8 · 2017-10-10, 16:09

That run searched for leading coefficients between 1 and 8000. Look at the LCmin and LCmax variables in the script. Or put them into the .n file as follows:

Code:

lcmin: 1
lcmax: 8000

LCstep will be ignored in a GPU run, it's only useful if you have several systems searching for a poly by CPU.

The time taken to search varies with the leading coefficient. Searching from 8000 to 16000 should take less time that 1 to 8000. So you can search a wider range if the coefficients are larger.

What range of leading coefficients to search for a given size of number is another issue. I've not had enough experience to offer advice.

Chris

2017-10-09, 03:00	#1
jacky Sep 2017 1001₂ Posts	Weird results of Msieve I have tested to factor C100 on two machines with GPU: USE_CUDA = True Command:"..\factMsieve.py example.n" 1、CPU:i7 3.60GHz RAM：8G GTX 745 elapsed time 7hours 2、CPU:E5-2670 2.60GHz RAM:64G Tesla P100-PCIE-16GB elapsed time 16hours The result confuesd me.It seems that GPU card doesn't work.Why Tesla is much slower? And if only test the first step,the command "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y" -t 4 -s gpu0polyfile"got the same result of "./msieve -np1 -nps "stage1norm=xeyy stage2norm=uevv x,y"-g 0 -t 8 -s gpu0polyfile".It seems both GPU and multi threads do not work.How to choose the number of thread?(-t number) Last fiddled with by jacky on 2017-10-09 at 03:01

2017-10-10, 16:09	#11
chris2be8 Sep 2009 81E₁₆ Posts	That run searched for leading coefficients between 1 and 8000. Look at the LCmin and LCmax variables in the script. Or put them into the .n file as follows: Code: lcmin: 1 lcmax: 8000 LCstep will be ignored in a GPU run, it's only useful if you have several systems searching for a poly by CPU. The time taken to search varies with the leading coefficient. Searching from 8000 to 16000 should take less time that 1 to 8000. So you can search a wider range if the coefficients are larger. What range of leading coefficients to search for a given size of number is another issue. I've not had enough experience to offer advice. Chris

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Weird	Dubslow	YAFU	14	2016-01-06 19:34
This is weird... Isn't it?	guido72	PrimeNet	18	2015-06-11 16:18
Weird factors	rekcahx	Miscellaneous Math	3	2011-11-01 23:25
weird abbreviations	science_man_88	Lounge	35	2010-11-28 04:56
something very weird	ixfd64	PrimeNet	1	2008-10-16 18:19

2017-10-09, 04:18	#5
VBCurtis "Curtis" Feb 2005 Riverside, CA 4,861 Posts	I don't think a GPU is any faster than CPU for a 100-digit input. That job is so small that invoking the GPU at all is a waste of effort; you may not be able to tell which is faster when the GPU works for something like 2 minutes. GPU-enhanced poly select doesn't become notably more powerful than CPU until 140 digits or so.

2017-10-09, 07:27	#7
jacky Sep 2017 3² Posts	Now Tesla 100 is 4 times faster than GTX 745.It seems not to have achieved the desired result.I think Tesla should be much faster.Should I give more threads?