search

NVIDIA / k8s-dra-driver

Dynamic Resource Allocation (DRA) for NVIDIA GPUs in Kubernetes
Apache License 2.0
195 stars 36 forks source link

Fix the cuda benchmark parameter value in the examples #129

Open yuanchen8911 opened 1 month ago

yuanchen8911 commented 1 month ago

  1. The number of bodies in the cuda benchmark should be a multiple of 256. Replace the existing values of 4226000 with 4226048 in the examples.

    > Compute 8.0 CUDA device: [NVIDIA A100-SXM4-40GB MIG 1g.5gb]
Warning: "number of bodies" specified 4226000 is not a multiple of 256.
Rounding up to the nearest multiple: 4226048.

  2. Remove mig+mps/sharing-demo/sharing-demo-job-1.yaml, which is not used.