[NVIDIA launches AI factory operation and maintenance platform Minerva that can analyze billions of data points every hour] On October 22, NVIDIA launched Minerva, an AI operation and maintenance platform, which aims to detect faults in advance, improve GPU runtime and improve AI factory output efficiency by integrating infrastructure and AI workload telemetry data. The platform can analyze billions of data points every hour, identify anomalies, predict potential failures, and consolidate associated alarms into actionable troubleshooting items. Minerva is also equipped with a native AI assistant that can query cluster health status and provide repair suggestions through natural language. Nvidia said that IREN is already using the platform, and Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI and Tesla plan to deploy it; Dell, Lenovo and Supermicro will also incorporate Minerva into their respective AI factory operation and maintenance product portfolios.