SALT LAKE CITY -- NVIDIA announced at KubeCon + CloudNativeCon North America 2026 three new open source components of its NVIDIA DSX AI Factory Platform: NVIDIA Kubernetes Engine (NVKE), NVIDIA Kubernetes Recipes (NVKR) and NVIDIA Snapshot. The tools are designed to help teams deploy, manage and scale GPU infrastructure on Kubernetes for AI factories, which need to offer customers managed Kubernetes services while retaining control of their infrastructure.
Managed Kubernetes Building Blocks for GPU Infrastructure #
NVKE and NVKR, now in public preview, provide open source building blocks for delivering managed Kubernetes services on GPU infrastructure. They automate cluster lifecycle management and provide signed, version-pinned images. Mirantis, Netris, OpenNebula Systems, Rafay, Spectro Cloud and vCluster, along with Firmus and SK Telecom, have aligned with NVIDIA on plans to bring NVKE and NVKR capabilities into their offerings.
Snapshot Cuts Inference Startup Times #
NVIDIA Snapshot targets the delay of bringing additional inference capacity online. The modular component restores checkpointed replicas in seconds without model reload or warmup, supporting faster startup, recovery and elasticity for inference and agentic workloads on Kubernetes. Design partner Crusoe, an upstream contributor to Snapshot, tested GLM-5.3 on Crusoe Cloud and cut the time for the workload to become ready from roughly eight minutes to 31 seconds. Saturn Cloud benchmarked Snapshot for scale-from-zero inference and reduced vLLM startup times from between two and 8.5 minutes to between six and 34 seconds.
NVIDIA is also contributing to Agent Substrate, the Google-originated CNCF Sandbox project for running long-running agent workloads on Kubernetes, and is ensuring it works with NVIDIA Dynamo for model inference and KAI Scheduler for shared cluster resource allocation. NVIDIA and Google are collaborating on an integration of Agent Substrate with OpenShell, NVIDIA’s open source agent security runtime, which NVIDIA is contributing to the CNCF.