Scope

This document describes how to deploy, configure, and operate NVIDIA Network Operator on Kubernetes clusters running on NVIDIA H100-based DGX SuperPOD and BasePOD systems. It is intended for architects, system administrators, and platform engineers using NVIDIA Base Command Manager (BCM) 11.x with NVIDIA Mission Control 2.x or later to support AI training, inferencing, and other data-intensive workloads.

The scope includes key concepts, an installation workflow using BCM, examples of required custom resources for secondary networks and SR-IOV-based networking, validation procedures, and troubleshooting guidance.

Hardware design, storage architecture, OS kernel, drivers, generic Kubernetes networking topics, and procedures for non-DGX SuperPOD/BasePOD environments are outside the scope of this document.