NVIDIA introduced Minerva on Oct. 22, a platform that pairs AI job information with hardware and network data to help operators investigate problems at AI factories.
Its data inputs span GPUs, networks, storage systems and schedulers, as well as the AI workloads running on them. NVIDIA says Minerva uses machine learning to flag unusual activity and connect alerts that may point to related issues.
An integrated assistant can draw on live telemetry to help teams review cluster health and consider possible fixes, the company said. Operators can make changes themselves or connect the platform to automation systems.
NVIDIA’s Mission Control release notes list Minerva 1.0.0 as pre-release beta software available through a limited engagement. The notes provide no general availability date.
IREN is already using Minerva. Aolani, Bitdeer AI, IBM, Jio, Lambda, Quake AI and Tesla plan to deploy it, NVIDIA said. IREN Chief Technology Officer Denis Skrinnikoff said the platform helps the company’s operations staff prioritize preventive maintenance as its AI factories expand.
Dell Technologies, Lenovo and Supermicro also plan to offer Minerva through their AI factory operations portfolios, according to NVIDIA.
