Prerequisites#

This guide assumes you have the following prerequisites in place:

Cert-Manager
Label nodes with intel.feature.node.kubernetes.io/gpu=true to indicate that they have Intel GPUs. You can do this using the following command:
```
kubectl label nodes <node-name> intel.feature.node.kubernetes.io/gpu=true
```

Intel GPU Operator#

Provides support for Intel GPUs in kubernetes clusters.

Installation#

Create the following directory structure for Intel GPU Operator:

intel-gpu-operator/
├── intel-device-operator.yml
├── intel-plugin-operator.yml
└── intel-plugin-repo.yml

Add the following content to intel-gpu-operator/intel-device-operator.yml:

---
apiVersion: helm.toolkit.fluxcd.io/v2
kind: HelmRelease
metadata:
  name: device-plugin-operator
  namespace: gpu-operator
spec:
  interval: 24h
  chart:
    spec:
      chart: intel-device-plugins-operator
      version: "0.35.0"
      sourceRef:
        kind: HelmRepository
        name: intel
        namespace: flux-system
      interval: 24h
  install:
    remediation:
      retries: 3
  upgrade:
    remediation:
      retries: 3

Add the following content to `intel-gpu-operator/intel-plugin-operator.yml

---
apiVersion: helm.toolkit.fluxcd.io/v2
kind: HelmRelease
metadata:
  name: gpu-device-plugin
  namespace: gpu-operator
spec:
  interval: 6h
  chart:
    spec:
      chart: intel-device-plugins-gpu
      version: "0.35.0"
      sourceRef:
        kind: HelmRepository
        name: intel
        namespace: flux-system
      interval: 6h
  install:
    remediation:
      retries: 3
  upgrade:
    remediation:
      retries: 3
  values:
    sharedDevNum: 4
    nodeFeatureRule: false
    nodeSelector:
      intel.feature.node.kubernetes.io/gpu: 'true'

Add the following content to `intel-gpu-operator/intel-plugin-repo.yml

---
apiVersion: source.toolkit.fluxcd.io/v1
kind: HelmRepository
metadata:
  name: intel
  namespace: flux-system
spec:
  interval: 6h
  url: https://intel.github.io/helm-charts

Commit and push the changes to your Git repository.