5Stack

Appearance

NVIDIA GPU Setup ​

Game streaming on NVIDIA GPUs needs three things on every node that will run a streamer pod:

  1. The NVIDIA driver and CUDA userland.
  2. The NVIDIA Container Toolkit (so containerd can expose GPUs into pods).
  3. The Kubernetes nvidia overlay applied to your cluster (registers the device plugin so pods can request nvidia.com/gpu).

1. Add the NVIDIA Container Toolkit repository ​

Before installing the drivers, install the prerequisites and add NVIDIA's apt repository so the container toolkit packages are available.

bash
sudo apt-get update && sudo apt-get install -y --no-install-recommends \
    ca-certificates \
    curl \
    gnupg2

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && 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

2. Install the NVIDIA drivers ​

On Ubuntu / Debian-based hosts:

bash
sudo ubuntu-drivers autoinstall

This picks the recommended driver branch for your GPU automatically. If you'd rather pin a specific branch, see NVIDIA's installation guide.

Reboot, then verify:

bash
nvidia-smi

You should see your GPU(s) listed with the driver version.

Secure Boot

If Secure Boot is enabled, nvidia-smi won't load the driver because the kernel module isn't signed by a trusted key. Enroll NVIDIA's Machine Owner Key so the module is accepted:

bash
sudo mokutil --import /var/lib/shim-signed/mok/MOK.der

You'll be prompted to set a one-time password. Reboot, complete MOK enrollment in the blue MOK Manager screen using that password, then re-run nvidia-smi. Alternatively, disable Secure Boot in your firmware.

3. Install the NVIDIA runtime ​

Install the container runtime packages so containerd can launch GPU-enabled containers:

bash
sudo apt install -y \
    nvidia-container-toolkit \
    nvidia-container-runtime

4. Restart K3s containerd ​

K3s detects the NVIDIA container runtime when it starts and adds the matching runtime entries to its generated containerd config. Restart the K3s agent on each GPU node:

bash
sudo systemctl restart k3s-agent

Confirm the runtime was detected:

bash
grep nvidia /var/lib/rancher/k3s/agent/etc/containerd/config.toml

You can now continue to the game streaming setup.