Go Back > Great Internet Mersenne Prime Search > Hardware > GPU Computing

Thread Tools
Old 2014-05-11, 09:19   #1
Feb 2014

5 Posts
Default cuFFT on multiple GPUs

Hi everybody,

the new CUDA6.0 Toolkit supports the use of 2 GPUs to perform FFT whose calculations can be distributed accross 2 GPUs under some boundary conditions. One of these preconditions (which is at the same time a restriction) is:

> the GPUs must both be on the same GPU board (e. g. GeForce GTX 690)

Is there anybody who has already tried to change / adapt the CUDALucas 2.05 source code to increase the performance when using CUDALucas on a Nvidia GeForce GTX 690?

HHfromG is offline   Reply With Quote
Old 2014-05-11, 18:39   #2
kracker's Avatar
"Mr. Meeseeks"
Jan 2012
California, USA

215910 Posts

I think running one "instance"per core will always be faster than one per multiple cores.
kracker is offline   Reply With Quote
Old 2014-05-11, 19:57   #3
xilman's Avatar
May 2003
Down not across

22·3·7·112 Posts

Originally Posted by kracker View Post
I think running one "instance"per core will always be faster than one per multiple cores.
Faster by what metric ---- latency or throughput?

Yes, that was a rhetorical question. The reason why I asked it should be obvious.
xilman is offline   Reply With Quote

Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
CUDALucas / cuFFT Performance on CUDA 7 / 7.5 / 8 Brain GPU Computing 13 2016-02-19 15:53
Anyone using GPUs to do DC, LL or P-1 work? chalsall GPU to 72 56 2014-04-24 02:36
GPUs impact on TF petrw1 GPU Computing 0 2013-01-06 03:23
GPUs in action? davieddy PrimeNet 21 2011-04-02 12:33
Multiple systems/multiple CPUs. Best configuration? BillW Software 1 2003-01-21 20:11

All times are UTC. The time now is 15:52.

Thu Aug 6 15:52:34 UTC 2020 up 20 days, 11:39, 1 user, load averages: 2.01, 2.00, 1.84

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.