This is a simple repository for preparing a k3s + nvidia/cuda base image that enables a K3d cluster to have access to your host machine's NVIDIA, CUDA-capable GPU(s).
Access to GitHub and GitHub Container Registry. Please follow the GitHub Container Registry instructions.
Docker and all of its dependencies must be installed. To build and push a new version of the Docker image to the GHCR, Docker Buildx must be installed and configured correctly.
For the container GPU test, a NVIDIA GPU with CUDA cores and drivers must be present. Additionally, the CUDA toolkit and NVIDIA container toolkit must be installed.
For Kubernetes testing and pre-requisites, please see Kubernetes Deployment for details.
Run the ./build.sh script to build and push a new image. Please see script for arguments and options.
Follow the instructions in the zarf-package-k3d-airgap repository for bootstrapping a K3d cluster that can access your NVIDIA GPUs.