40,947
edits
Hyper-Q / MPS/fr: Difference between revisions
Jump to navigation
Jump to search
Updating to match new version of source page
(Created page with "Vous pouvez alors utiliser la fonctionnalité MPS si vous avez plus d'un fil CPU qui a accjs au GPU. Ceci se produit quand vous exécutez une application hybride MPI/CUDA, une application hybride OpenMP/CUDA application ou plusieurs applications séquentielles CUDA (<i>GPU farming</i>).") |
(Updating to match new version of source page) |
||
| Line 55: | Line 55: | ||
<div lang="en" dir="ltr" class="mw-content-ltr"> | <div lang="en" dir="ltr" class="mw-content-ltr"> | ||
In the above example, we share a single V100 GPU between 8 instances of <code>my_code</code> (which takes a single argument-- the loop index $i). We request 8 CPU cores (#SBATCH -c 8) so there is one CPU core per application instance. The two important elements are <code>&</code> on the code execution line, which sends the code processes to the background, and the <code>wait</code> command at the end of the script, which ensures that the job runs until all background processes end. | In the above example, we share a single V100 GPU between 8 instances of <code>my_code</code> (which takes a single argument-- the loop index $i). We request 8 CPU cores (#SBATCH -c 8) so there is one CPU core per application instance. The two important elements are | ||
* <code>&</code> on the code execution line, which sends the code processes to the background, and | |||
* the <code>wait</code> command at the end of the script, which ensures that the job runs until all background processes end. | |||
</div> | </div> | ||
[[Category:Software]] | [[Category:Software]] | ||