Quake AI to deploy NVIDIA Minerva for AI factory ops

Quake AI's planned deployment of NVIDIA Minerva builds on an existing relationship: RUM Group CEO Chris Pavlovski noted that NVIDIA has powered Quake AI's infrastructure for years, with more than 22,000 GPUs in operation today and NVIDIA Vera Rubin deployments planned for 2027. That scale explains why operational visibility has become a priority. Bare-metal customers want direct control over their compute, but reliable performance at scale requires knowing how the underlying infrastructure is behaving. Minerva is designed to ingest and correlate telemetry from GPUs, networking, storage, schedulers and AI workloads, helping operators detect anomalies and anticipate failures before they disrupt training or inference. By moving from reactive troubleshooting toward proactive maintenance, Quake AI aims to support the uptime and performance its customers expect. The deployment, once completed, would pair direct access to compute with the operational discipline needed for dependable performance at scale.