nvidia-gpu-operator.yaml

# NVIDIA GPU Operator ArgoCD Application.
# Required for GPU nodes to advertise nvidia.com/gpu as an allocatable
# resource. Without this, GPU-requiring pods stay Pending.
#
# NOT NEEDED ON AWS. Since NIC v0.9.0, marking a node group `gpu: true`
# in your NIC config makes NIC install the GPU Operator itself during
# `nic deploy`, into the `gpu-operator` namespace. Applying this
# alongside it gives you two releases fighting over the same DaemonSets.
# Use this only on providers NIC does not cover (k3s, on-prem, GCP,
# Azure).
#
# Use this values block when your GPU nodes boot from an AMI that ships BOTH
# the NVIDIA driver and the container toolkit: it disables both so the
# operator only adds the device plugin. If your AMI ships the driver but not
# the toolkit (e.g. NIC's AL2023 NVIDIA AMI), keep toolkit enabled instead
# (see the Installation guide's nvidia-gpu-operator step).
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: nvidia-gpu-operator
  namespace: argocd
  annotations:
    argocd.argoproj.io/sync-wave: "2"
  finalizers:
    - resources-finalizer.argocd.argoproj.io
spec:
  project: nebari-apps

  source:
    chart: gpu-operator
    repoURL: https://helm.ngc.nvidia.com/nvidia
    targetRevision: v25.10.1
    helm:
      releaseName: nvidia-gpu-operator
      values: |
        # Disable driver and toolkit when the AMI ships both. Remove these
        # overrides if your nodes need the operator to install the driver.
        driver:
          enabled: false
        toolkit:
          enabled: false

  destination:
    server: https://kubernetes.default.svc
    namespace: gpu-operator

  syncPolicy:
    automated:
      prune: true
      selfHeal: true
    syncOptions:
      - CreateNamespace=true
      - ServerSideApply=true
    retry:
      limit: 5
      backoff:
        duration: 5s
        factor: 2
        maxDuration: 3m