Message boards : GPUs : Ability to stop CPU work while GPU is working and vice versa.
Message board moderation
Author | Message |
---|---|
Send message Joined: 29 Jan 15 Posts: 18 |
Hello everyone, Apologies is this question was asked before, but is there a way to set up BOINC to first run the GPU and CPU task separately after one or the other finishes? I have a laptop and if I run both CPU and GPU together it overheats. Currently I have been suspending the cpu project and then the GPU project, but is there a better way to handle it? Thank you for your responses |
Send message Joined: 5 Oct 06 Posts: 5121 |
This is very similar to github #2214 - perm 'one at a time' from {CPU}, {nvidia GPU}, {intel gpu}. We've never found a simple, universal way of controlling temperatures of (especially) GPU devices from every manufacturer under every operating system, or to handle different usage preferences at different times of day or alongside different workloads. Sadly, I don't think it's likely to happen in the near future, either, unless a new volunteer comes along and has a lightbulb moment. |
Send message Joined: 29 Jan 15 Posts: 18 |
Hi Richard, Thank you for replying, I will take a look at that thread to come up with some ideas. I guess I will just have to keep doing what I am doing now, manually letting one project get all the Gpu one week and the other one get all the CPU time next week. Perhaps one way to do it is maybe create a program or dummy executable that starts up when GPU work is running. I know that in the preferences options there is a way to let a program stop all BOINC functions if it detects that it starts. This could have the CPUs put on hold, but let the GPU work units continue, although I am not sure if it would be possible to have that all in one BOINC app. Seems like that would require a seperate program for CPU BOINC and GPU BOINC. This would also mean that there has to be some way to limit the number of GPU work units that boinc would pull on a daily basis (aka max 10 units per project) so that it would not just keep running and have all the CPU work stop. |
Send message Joined: 5 Mar 08 Posts: 272 |
Just setup app_info or app_config with the number of CPUs in the system. When a GPU app runs it will reserve all CPUs. You’ll need one for each project you run. See app_config for details on setting up an app_config MarkJ |
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.