Build a Validated Kubernetes Runtime for NVIDIA Dynamo
NVIDIA, Friday, October 30th, 2026
NVIDIA describes AI Cluster Runtime (AICR), which packages validated Kubernetes recipes for deploying the Dynamo distributed inference framework.
The DSX AI Factory Platform uses AICR to capture runtime needs such as GPU driver stacks, Dynamic Resource Allocation, gang scheduling, node health signals and Prometheus metrics in reproducible recipes with pinned versions.
Dynamo 1.4.2 uses TCP for the request plane and ZMQ for KV-cache events, supporting cache-aware routing through either its native Frontend router or the Gateway API Inference Extension and Endpoint Picker path.
AICR follows a five-step flow: snapshot the cluster, generate a recipe, render a deployment bundle, deploy with Helmfile or GitOps, and validate the runtime before installing applications.