NVIDIA Open-Sources GPU Cluster Management Plane NVKE

On November 13, 2026, at 17:00, the AI section of DEV Community reported that NVIDIA has open-sourced NVKE, a management plane for GPU infrastructure operators. According to the report, NVKE lets an operator create and run Kubernetes clusters for its own customers directly on bare metal. Each workload cluster is given a dedicated Kubernetes control plane, its own etcd, and its own set of node pools, while the operator is responsible for running the management plane that provisions them.

The report also outlines how the system is brought up and configured. A throwaway seed cluster is used to stand up the management cluster, and the management cluster then provisions workload clusters on request through the NVKE API. In this way, NVKE works by using the bootstrap cluster to dynamically configure the workload clusters it manages, with the control plane role handled within that structure.

At this point, this is the only publicly available report on the release, so there is no additional verifiable information yet about further NVKE versions, documentation, or deployment progress. The release time, the project name NVKE, and the described architecture and provisioning flow rest solely on this single source.