Documentation

The DRA Driver for NVIDIA GPUs is a Kubernetes Dynamic Resource Allocation driver that lets workloads request GPUs, MIG slices, and VFIO passthrough devices (gpu.nvidia.com), and provisions ephemeral Multi-Node NVLink fabrics via IMEX (compute-domain.nvidia.com) so pods on different nodes can share GPU memory over NVLink. It targets Kubernetes 1.32+ with DRA enabled.

• architecture.md — components and request flows.
• development.md — build, test, and local-cluster loop.