GPU Support for Kubernetes Workloads

GPU Support

Step 1: Enable GPU support

Intel GPU (iGPU for transcoding)

No host prerequisites needed - the Intel GPU Plugin DaemonSet handles everything.

The Intel GPU plugin manifest lives in infrastructure/gpu/intel-plugin.yaml.

Verify after deployment:

kubectl describe node | grep gpu.intel.com/i915

NVIDIA GPU

Install NVIDIA Container Toolkit

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | \
  sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg --yes
curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
  sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list
sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit

Configure Containerd for NVIDIA

These three commands are required for NVIDIA to work in Kubernetes.

sudo nvidia-ctk runtime configure --runtime=containerd --set-as-default --cdi.enabled
sudo nvidia-ctk cdi generate --output=/etc/cdi/nvidia.yaml
sudo systemctl restart containerd

Verify after deployment:

kubectl describe node | grep nvidia.com/gpu

The NVIDIA GPU plugin manifest lives in infrastructure/gpu/nvidia-plugin.yaml.