Radar / AI infrastructure / NVKE
NVKE: NVIDIA's open-source management plane for GPU Kubernetes on bare metal
NVIDIA open-sourced NVKE, a preview management plane that runs isolated Kubernetes clusters with their own control plane and etcd on bare-metal NVIDIA GPU fleets.
Why it matters
Turns owned GPU hardware into a managed-Kubernetes service: each workload cluster gets its own control plane, TPM-attested nodes via SPIFFE/SPIRE, and per-cluster etcd — without a cloud provider.
What you could build with it
A GPU-cloud startup with its own hardware could trial NVKE's quickstart to offer per-tenant Kubernetes clusters on bare metal, comparing provisioning time and isolation against their existing Cluster API setup.
Does it hold up?
Preview only, and NVIDIA says not for production — no management-plane upgrades or backup/restore yet, and real installs need NICo access; the local quickstart on simulated machines is the practical way to evaluate it today.
Built with NVKE
- Nvidia previews NVKE, a bare-metal GPU-Kubernetes factoryaiweekly.co · Breakdown of the preview: the self-hosting factory model, TPM attestation, and what NVKE explicitly is not.
- NVIDIA Kubernetes Engine Enters Previewnews.lavx.hu · Reports the Nov 6 preview release, the NICo-only provider limitation, and the companion NVKR project.
- dsx-ai-factory/nvkrgithub · 16 ★ · Companion project that builds and signs the node OS images and NVKE components that NVKE deploys.
Learn more
First spotted on github: source.
More AI infrastructure
colibri: run 744B-parameter MoE models on consumer hardwarePure-C inference engine with zero dependencies that streams MoE experts from NVMe disk, keeping hot experts on GPU…infra · JEV 0.78claude-mem: persistent memory plugin that lets coding agents pick up where the last session left offPlugin that hooks five Claude Code lifecycle events, compresses session activity into a local SQLite + Chroma store…infra · JEV 0.74FreeToken: datacenter-scale MoE serving on desktop GPUsAn edge-native mixture-of-experts serving engine from UC Berkeley researchers that runs 290B+ parameter MoE models on…infra · JEV 0.74XLA Serving Engine: Google open-sources its C++ inference runtimeGoogle's general-purpose model export and inference runtime for the OpenXLA ecosystem, now open source under Apache 2.0.infra · JEV 0.67TorchTPU: Google's native PyTorch stack for TPUs, now open sourceAnnounced at PyTorch Conference 2026 on October 20: Google open-sourced TorchTPU, its native PyTorch backend for Cloud…infra · JEV 0.66billion-context: month-long agent sessions in a 100K context windowA local proxy that sits between a coding agent and its model API, letting the model fold old messages into searchable…infra · JEV 0.65
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.