Validate GPU Cluster Readiness Before AI Workloads Land

NVIDIA Cluster Readiness Engine (NVCRE) is an open‑source Kubernetes controller that runs real distributed workloads across topology‑aware node groups to certify GPU cluster readiness before production AI jobs are launched. By using custom resources for Certification, Workflow and Job, it attributes failures to specific nodes or categories (e.g., NCCL communication, NeMo pretraining) and applies adaptive fault isolation to narrow down problematic elements. NVCRE integrates with NVIDIA AI Cluster Runtime and NVSentinel for continuous health telemetry, forming part of the NVIDIA DSX OS operating layer, and can be triggered via its CLI (nvcrectl) to run end‑to‑end validation tests on any GPU cluster.

Source: NVIDIA Developer | 2026-09-23