Allocate a part of device memory by percentage to container

To allocate a certain size of GPU device memory by percentage, you need only to assign nvidia.com/gpumem-percentage besides nvidia.com/gpu.

apiVersion: v1
kind: Pod
metadata:
  name: gpu-pod
spec:
  containers:
    - name: ubuntu-container
      image: ubuntu:22.04
      command: ["bash", "-c", "sleep 86400"]
      resources:
        limits:
          nvidia.com/gpu: 2 # requesting 2 vGPUs
          nvidia.com/gpumem-percentage: 50 # each vGPU requests 50% of device memory

note

Prefer specifying memory using one of these resources. If both are set and nvidia.com/gpumem is a positive integer, it takes precedence when determining the memory request. See Allocate device memory for details.