Nvidia launched Minerva on Thursday, an AIOps platform that pulls telemetry from AI infrastructure and workloads into one system and uses purpose-built AI models to flag problems before they stall a cluster. Nvidia is pitching it as a way to raise "AI factory yield," which it measures in GPU hours, goodput, cluster uptime and token production.
The problem Nvidia is targeting is small faults with large consequences. A single degraded component, such as an overheating GPU or an underperforming NVMe drive, can slow an entire job and is easy to miss among thousands of machines.
Minerva collects signals across Nvidia GPUs, networking, storage, schedulers and workloads, then builds a dynamic knowledge graph that maps GPUs, SuperNICs, switches and cables to the jobs running on them. Operators can use that map to trace a slow job back to the responsible part, even in clusters with hundreds of thousands of GPUs.
Key facts
- 01Nvidia launched Minerva, an AIOps platform for AI factory observability, on October 22, 2026.
- 02Minerva analyzes billions of data points per hour using supervised and unsupervised machine learning.
- 03IREN is an early adopter; Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI and Tesla plan to deploy it.
- 04Dell Technologies, Lenovo and Supermicro will offer Minerva in their AI factory operations portfolios.
- 05In Nvidia's internal testing, Minerva cut repair time for an NVMe power fault from multiple days to six hours.
The platform analyzes billions of data points per hour with supervised and unsupervised machine learning. It learns what normal behavior looks like, detects anomalies it has not seen before, predicts failures and groups related events into single incidents with a likely root cause.
Minerva also ships with a native AI assistant that answers natural-language questions grounded in live telemetry and recommends fixes. Teams can apply those fixes by hand or automate them through infrastructure APIs and external agentic AI tools.
Bitdeer AI, which adopted Minerva the same day, said the assistant is called Sherlock and that an MCP server supports natural-language troubleshooting. Bitdeer AI also said the software integrates with Nvidia DSX OS and can run standalone.
Bitdeer AI cited results from Nvidia's internal testing: Minerva cut mean time to repair for an NVMe power fault from multiple days to six hours, and it consolidated 81 noisy events into one alert, saving 54 hours of cluster uptime.
IREN is an early adopter. Denis Skrinnikoff, IREN's chief technology officer, said maximizing GPU productivity is central to serving its customers as it scales its AI factories, and that Minerva lets his team spend more time on proactive maintenance than on troubleshooting.
“NVIDIA Minerva gives our operations team one view across the network, the GPUs and the jobs running on them”— Denis Skrinnikoff, Chief Technology Officer, IREN
Nvidia said Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI and Tesla plan to deploy Minerva, while Dell Technologies, Lenovo and Supermicro will offer it as part of their AI factory operations portfolios. Lenovo said Minerva fits its Hybrid AI Factory with Nvidia offering.
Nvidia did not disclose pricing, regional availability or a detailed deployment timeline for Minerva.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
