Update container_cluster: add MPS strategy on gpu_sharing block (#10471…

…) (#17985) [upstream:552d835ed5ba412770659dc926e07c5625ca1506] Signed-off-by: Modular Magician <[email protected]>
hashicorp · Apr 29, 2024 · 5e93d19 · 5e93d19
1 parent da580e3
commit 5e93d19
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/website/docs/r/container_cluster.html.markdown b/website/docs/r/container_cluster.html.markdown
@@ -1032,6 +1032,7 @@ sole_tenant_config {
 * `gpu_sharing_strategy` (Required) - The type of GPU sharing strategy to enable on the GPU node.
     Accepted values are:
     * `"TIME_SHARING"`: Allow multiple containers to have [time-shared](https://cloud.google.com/kubernetes-engine/docs/concepts/timesharing-gpus) access to a single GPU device.
+    * `"MPS"`: Enable co-operative multi-process CUDA workloads to run concurrently on a single GPU device with [MPS](https://cloud.google.com/kubernetes-engine/docs/how-to/nvidia-mps-gpus)
 
 * `max_shared_clients_per_gpu` (Required) - The maximum number of containers that can share a GPU.