mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   GMP-ECM (https://www.mersenneforum.org/forumdisplay.php?f=55)
-   -   ECM for CUDA GPUs in latest GMP-ECM ? (https://www.mersenneforum.org/showthread.php?t=16480)

Batalov 2012-02-11 00:23

For Windows, it would have been nice if someone could downgrade the build/solution files vs2010 to vs2008 and post. For comparison, CUDA toolkit contains scripts for 2005, 2008, 2010 -- that is kinda user friendly. Not everyone has 2010. (Well, temporarily one can get the trial license.)

xilman 2012-02-11 00:27

[QUOTE=Batalov;288969]For Windows, it would have been nice if someone could downgrade the build/solution files vs2010 to vs2008 and post. For comparison, CUDA toolkit contains scripts for 2005, 2008, 2010 -- that is kinda user friendly. Not everyone has 2010. (Well, temporarily one can get the trial license.)[/QUOTE]If you want it, why don't you do it?

Prime95 2012-02-11 01:21

[QUOTE=Batalov;288969]Not everyone has 2010. (Well, temporarily one can get the trial license.)[/QUOTE]

VS 2010 is free (the express edition). Yes, it is 32-bit only, but you can get the Windows SDK for free which contains the 64-bit compiler.

Batalov 2012-02-11 02:51

[QUOTE=xilman;288971]If you want it, why don't you do it?[/QUOTE]
Because it is ugly! I am not talking about .sln files; for the .vcxproj -> .vcproj conversion most of the internet based advices amount to 'you might be best served by using the "New project from existing code" wizard to build a new VC2008 project for the code rather than trying to convert the existing project.' It is best done by the authors who know their source and dependencies.

I'll try the free version.

frmky 2012-02-11 08:03

[QUOTE=R.D. Silverman;288966]Is the code specific to a particular GPU? How portable is it?[/QUOTE]
It uses CUDA, thus requires an nVidia GPU supporting CC 1.3 or higher.

frmky 2012-02-11 08:09

[QUOTE=xilman;288906]Each stage one took 70 seconds on a GTX-460.[/QUOTE]
The GTX 460 has 7 MP's. Did you do 224 curves in parallel or more? Are there memory latencies that are hidden by doing more?

debrouxl 2012-02-11 08:59

On my GT540M (admittedly a fairly low-end model, with 2 MPs), under Debian unstable x86_64, gpu_ecm with 64 curves in parallel seems to be somewhat slower than CPU-based stage 1 (tuned GMP-ECM binary) running on Core i7-2670QM @ 2.2 GHz.
I have tested B1 bounds from 5e4 to 16e6, and 32, 64 or 128 parallel curves. 32 curves has throughput markedly slower than 64, but 128 is hardly better than 64 for throughput.

xilman 2012-02-11 09:14

[QUOTE=frmky;289014]The GTX 460 has 7 MP's. Did you do 224 curves in parallel or more? Are there memory latencies that are hidden by doing more?[/QUOTE]The first 896 curves were done in four batches of 224 --- the default. The second ran all in parallel with a block of 32x32 and grid of 70x1x1. I seem to have lost the detailed timing information for the earlier curves :sad: As best I recall, running 896 took slightly less than running 224 four times but I could be quite wrong.

xilman 2012-02-11 09:23

Big oops.
 
I screwed up computing the time per curve :redface:
[QUOTE=xilman;288906]Each stage one took 70 seconds on a GTX-460. The latest ECM takes 679 seconds per stage 1 on a single core of a 1090T clocked at 3264MHz, so the GPU version is close to 10 times faster in this situation.[/QUOTE]

1792 curves took 141 hours to run. I evaluated (1792 * 141 / 3600) to obtain the quoted figure of 70 seconds per curve.

The correct expression is (141 * 3600 / 1792), which evaluates to 283 seconds per curve.
Although this is four times worse than the initial figure, it is still 2.4 times faster than a singe core.

Sorry about that.

Karl M Johnson 2012-02-11 11:23

[QUOTE=frmky;289013]It uses CUDA, thus requires an nVidia GPU supporting CC 1.3 or higher.[/QUOTE]
DP FP calculations are being performed ?

jasonp 2012-02-11 12:30

No, it looks to be integer only. Compute Capability 1.3 also has advanced synchronization primitives that the CC 1.3 branch uses. The code is [url="https://gforge.inria.fr/scm/viewvc.php/trunk/gpu/?root=ecm"]here[/url] for anyone who's interested.


All times are UTC. The time now is 19:10.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.