Support New AI Frameworks Faster Using NVIDIA Workload Map
According to NVIDIA Developer. NVIDIA introduces Workload Map to provide a common Kubernetes workload description for distributed AI frameworks.
- • NVIDIA Workload Map provides a common Kubernetes workload description for distributed workloads.
- • The project includes a catalog of prebuilt definitions for multiple frameworks and tools.
- • A single definition supports workload inspection, gang scheduling, and platform visibility.
Overview
NVIDIA Workload Map provides a common Kubernetes workload description that enables schedulers, controllers, and platforms to interpret distributed workloads without framework-specific parsing code. Operating an AI factory requires a shared understanding of distributed workloads including their components, pod templates, replica counts, and status.
Catalog and Integration
The project ships a catalog of prebuilt definitions for JobSet, PyTorchJob, RayCluster, RayJob, RayService, KServe, Knative, MPIJob, NVIDIA NIM, LeaderWorkerSet, Milvus, Grove PodCliqueSet, and NVIDIA Dynamo. Elotl's Nova uses Workload Map to interpret workload structure and resource requirements for multi-cluster scheduling decisions.
- 1. developer.nvidia.comOpen source