Thread 'Projects for GPUs with less than 256mb of memory'

Message boards : GPUs : Projects for GPUs with less than 256mb of memory
Message board moderation

To post messages, you must log in.

AuthorMessage
Luis Rodrigues

Send message
Joined: 23 Nov 08
Posts: 4
Spain
Message 37225 - Posted: 20 Mar 2011, 9:21:19 UTC

Hi,
Have 2 questions, please;
Are any projects for NVIDIA GPUs with less than 256mb of memory?
and how to limit/force/config those projects only to use the GPU, leaving another non-GPU projects (i.e. World Community Grid) using all CPU resources?
Thanks in advance
ID: 37225 · Report as offensive
ProfileJord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15560
Netherlands
Message 37228 - Posted: 20 Mar 2011, 11:12:01 UTC - in response to Message 37225.  
Last modified: 20 Mar 2011, 11:12:17 UTC

Are any projects for NVIDIA GPUs with less than 256mb of memory?

Collatz can use 128MB and less.

and how to limit/force/config those projects only to use the GPU, leaving another non-GPU projects (i.e. World Community Grid) using all CPU resources?

Technically, this isn't possible, since no GPU application runs on the GPU only. All GPU applications run on the CPU and use cycles of it. So you can't disable CPU use completely.

However, on the project's preferences page you will find options such as "Use CPU", "Use ATI GPU" and "Use Nvidia GPU". Depends on the project of course and whether or not they do have GPU applications and then which vendor's. There you can make your pick. By not enabling "Use CPU", you won't completely disable the CPU, you just won't get the CPU application and work for it from that project.
ID: 37228 · Report as offensive
Luis Rodrigues

Send message
Joined: 23 Nov 08
Posts: 4
Spain
Message 37230 - Posted: 20 Mar 2011, 12:01:55 UTC - in response to Message 37228.  

Hi Jord,
Thanks for the fast answer.
I have already Collatz working and I'll change the config to test it.
Thanks again.
ID: 37230 · Report as offensive

Message boards : GPUs : Projects for GPUs with less than 256mb of memory

Copyright © 2024 University of California.
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.