NVIDIA Details Minerva AIOps Platform for AI Factories
NVIDIA describes Minerva as an AI-native AIOps platform designed around AI factory topology and workloads. It collects telemetry across GPUs, networking, storage and workloads; maps infrastructure to active jobs in a dynamic knowledge graph; and uses topology-aware correlation to group dependent symptoms into incidents. The platform includes a native AI agent, Sherlock, that lets operators query live telemetry and the knowledge graph for root-cause analysis in natural language. NVIDIA says its testing showed reductions in notified events and tickets for supported scenarios. In one internal case, Minerva linked 81 events and anomalies to an affected job as a single incident, reducing repair time and saving about 54 hours of total cluster uptime; these results are vendor-reported. Minerva is also being aligned with DSX OS, with capabilities beginning with the data collector planned to become open-source parts of DSX OS.
Minerva is designed to address the operational complexity of large AI training environments, where faults can generate many disconnected alerts and require extensive cross-team triage. By linking infrastructure telemetry with workload context, grouping related symptoms and narrowing the search space for root-cause analysis, it aims to reduce time lost to diagnosis and repair.
AI Factories are some of the most technologically complex operational environments ever created. Hundreds of thousands of GPUs, NVIDIA Spectrum-X Ethernet scale-out, scale-across and scale-in fabrics, NVIDIA NVLink scale-up fabrics, distributed schedulers, and multi-week training runs operate in continuous lockstep.These layers function as one production system even when different teams own their telemetry. In synchronous distributed training, a job advances at the pace of its slowest rank. One degraded GPU, interface, storage device, or switch path can waste useful time across many healthy resources.
The failures that hurt yield are often subtle: A GPU runs a few degrees hotter than its peers. An NVMe device loses performance only under sustained load. A switch generates a pause-frame storm that looks like dozens of unrelated symptoms. None of these signals necessarily crosses a static threshold. Together, they create stragglers, extended step times, and delayed checkpoints.
Diagnosis becomes harder because the evidence is split across teams. Network operators see interface counters and fabric events. Compute operators see node and GPU health. ML teams see changing step times, failed jobs, or unstable inference latency. Each view is accurate, but the relationships between them remain hidden. Triage becomes a chain of dashboard searches and cross-team handoffs, which can turn a contained fault into hours or days of lost production time before the affected node or fabric path is identified.
#NVIDIA#NVIDIA Minerva#Sherlock#NVIDIA DSX OS#NVIDIA Spectrum-X Ethernet#NVIDIA NVLink