NVIDIA launches Minerva to detect problems across AI data centers
NVIDIA introduced Minerva, a platform that analyzes hardware and workload data to help operators identify failures and keep AI computing systems productive.
NVIDIA
NVIDIA launched Minerva on October 22 to help operators detect problems in AI data centers. The company said the platform combines operational data with AI models to identify faults and support maintenance.
Minerva maps connections between computing equipment and the jobs it supports. NVIDIA said it can analyze billions of measurements hourly, identify unusual behavior and combine related alerts into incidents with a probable cause. An included AI assistant lets operators investigate problems through conversation.
The software addresses a challenge in large computing clusters: one struggling component can slow a broader job. Connecting hardware conditions with workload performance is intended to help operators locate those bottlenecks. NVIDIA said teams can implement suggested fixes manually or connect the platform to automated systems.
NVIDIA named IREN as an existing user. Planned deployments include IBM, Jio, Lambda and Tesla, while Dell Technologies, Lenovo and Supermicro will offer Minerva within their services for operating AI infrastructure. The announcement did not specify deployment dates for those planned users.
The day's biggest AI news in one short email, every morning. Plus an instant heads-up when a reset or a big release lands.
More on NVIDIA
- Oct 22Nvidia B300 servers reportedly linked to China Mobile computing hubRumor
- Oct 22NVIDIA expands Cosmos tools for world model training and simulationOfficial
- Oct 22NVIDIA details retail AI projects spanning shopping assistants and warehouse simulationOfficial
- Oct 21Nvidia details Kumo Relational model and claims gains on database prediction testsOfficial
- Oct 21Nvidia’s financing role grows as AI infrastructure companies face funding pressureNews
- Oct 21Recursive releases Flint coding model for local hardwareOfficial