NVIDIA Fleet Intelligence

Type: Tool Tags: NVIDIA, Fleet Intelligence, GPU fleet, health monitoring, predictive failure, AI cloud, operations, NCX Related: NVIDIA-Cloud-Accelerator-NCX, NVIDIA-DGX-Cloud, NVIDIA-DCGM, NVIDIA-NVSentinel, NVIDIA-Project-GPUd, NVIDIA-Mission-Control Sources: https://docs.nvidia.com/ncx/index.html; https://docs.nvidia.com/fleet-intelligence/latest/index.html Last Updated: 2026-04-29

Summary

NVIDIA Fleet Intelligence is NVIDIA’s agent-based managed service for GPU fleet health monitoring and predictive failure signals. NCX positions it as an AI-cloud operations component for continuous GPU health monitoring, maximum uptime, and stability. The current public docs surface includes a Fleet Intelligence User Guide v1.1, release notes, dashboard, inventory, events/alerts, detailed debugging analysis, reports, settings, enrollment tokens, retention settings, alert configuration, and agent enrollment troubleshooting.

Detail

Fleet Intelligence belongs in the wiki as the managed fleet health layer in NVIDIA cloud partner and AI factory operations. It is distinct from low-level telemetry collection because its focus is fleet-level inventory, events, alerts, analysis, reporting, and predictive signals for operators responsible for many GPU systems.

The current docs hierarchy shows pages for machine details, events/alerts, standard groupings, tags, detailed debugging analysis, reports, settings, data retention, alert configuration, and agent enrollment troubleshooting. That makes it a natural companion to NVIDIA-DCGM, NVIDIA-NVSentinel, NVIDIA-Project-GPUd, and NVIDIA-Mission-Control rather than a replacement for them.

Connections

• NVIDIA-Cloud-Accelerator-NCX - NCX lists Fleet Intelligence among the AI cloud software components.
• NVIDIA-DGX-Cloud - Fleet Intelligence is relevant to cloud-scale GPU fleet uptime and stability.
• NVIDIA-DCGM - DCGM provides low-level GPU health and telemetry that fleet-level tools can consume.
• NVIDIA-NVSentinel - NVSentinel handles Kubernetes-native fault detection/remediation; Fleet Intelligence provides managed fleet insights.
• NVIDIA-Project-GPUd - GPUd is a lightweight GPU telemetry agent adjacent to fleet monitoring.
• NVIDIA-Mission-Control - Mission Control and Fleet Intelligence both target AI factory operations and reliability.

Source Excerpts

• “continuous GPU health monitoring and predictive failure signals”
• “NVIDIA Fleet Intelligence User Guide”