NVIDIA has introduced Minerva, an AI operations platform designed to enhance infrastructure reliability and maximize output efficiency for large-scale AI deployments. The system consolidates telemetry data from both hardware infrastructure and AI workloads, analyzing billions of data points per hour to detect anomalies, predict potential failures, and merge related alerts into actionable incident items. Minerva features a native AI assistant that allows users to query cluster health status in natural language and receive automated remediation suggestions. IREN is already utilizing the platform, while major industry players including Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI, and Tesla have announced plans to deploy it. Additionally, Dell, Lenovo, and Supermicro will integrate Minerva into their respective AI factory operations portfolios.