rsnt_translations
56,420
edits
Using GPUs with Slurm: Difference between revisions
Jump to navigation
Jump to search
no edit summary
No edit summary |
No edit summary |
||
| Line 19: | Line 19: | ||
The following form can also be used: | The following form can also be used: | ||
--gres=gpu[[:type]:number] | --gres=gpu[[:type]:number] | ||
This is older, and we expect it will no longer be supported in some future release of Slurm. We recommend that you replace it in your scripts with the above --gpus-per-node form. | This is older, and we expect it will no longer be supported in some future release of Slurm. We recommend that you replace it in your scripts with the above <code>--gpus-per-node</code> form. | ||
<!--T:61--> | <!--T:61--> | ||
There are a variety of other directives that you can use to request GPU resources: --gpus, --gpus-per-socket, --gpus-per-task, --mem-per-gpu, and --ntasks-per-gpu. Please see the Slurm documentation for [https://slurm.schedmd.com/sbatch.html sbatch] for more about these. Our staff did not test all the combinations of these; if you don't get the result you expect, [[Technical support|contact technical support]]. | There are a variety of other directives that you can use to request GPU resources: <code>--gpus</code>, <code>--gpus-per-socket</code>, <code>--gpus-per-task</code>, <code>--mem-per-gpu</code>, and <code>--ntasks-per-gpu</code>. Please see the Slurm documentation for [https://slurm.schedmd.com/sbatch.html sbatch] for more about these. Our staff did not test all the combinations of these; if you don't get the result you expect, [[Technical support|contact technical support]]. | ||
<!--T:15--> | <!--T:15--> | ||
| Line 63: | Line 63: | ||
<!--T:55--> | <!--T:55--> | ||
(*) | (*) <b>Compute Capability</br> is a technical term created by NVidia as a compact way to describe what hardware functions are available on some models of GPU and not on others. | ||
It is not a measure of performance. It is relevant only if you are compiling your own GPU programs. See the page on [[CUDA#.22Compute_Capability.22|CUDA programming]] for more. | It is not a measure of performance. It is relevant only if you are compiling your own GPU programs. See the page on [[CUDA#.22Compute_Capability.22|CUDA programming]] for more. | ||