NVIDIA Cloud Functions
A Kubernetes-based platform for deploying and routing GPU-accelerated inference, streaming, and batch workloads at scale, handling function lifecycle, invocation routing, GPU cluster integration, and observability.
NVIDIA Cloud Functions manages GPU-accelerated workloads as either long-running functions, invokable endpoints for inference or streaming, packaged as a container or a Helm chart when multiple coordinated services are needed, or as asynchronous tasks for run-to-completion work such as batch inference, evaluation, fine-tuning, or data preparation.
Its control plane exposes the NVCF API and manages function and deployment state and secrets, while its invocation plane receives HTTP, streaming, and gRPC requests and applies routing and rate limiting before sending work to running workloads. GPU clusters connect through a companion cluster agent, which registers GPU resources and manages workload execution on GPU nodes, and function artifacts are pulled from container or chart registries the deployment can reach.
This monorepo contains the service code, deployment assets, documentation, examples, CLI code, agent skills, and validation tooling, and is in the process of moving to a new GitHub organization, with old clone and module paths redirected during the transition.