NVIDIA AI Enterprise Infra 7.2 Release Notes#

Release Highlights#

RTX PRO Blackwell Support - vGPU for Compute 19.2 adds support for NVIDIA RTX PRO 6000 Blackwell Server Edition on VMware vSphere 9.0.1
Infrastructure Updates - GPU Data Center Driver 580.95.05, DOCA-OFED v25.7.0, and Container Toolkit v1.18.0 for enhanced performance
Kubernetes Orchestration - GPU Operator v25.10.0, Network Operator v25.7.0, DPU Operator v25.7.1, and NIM Operator v3.0.1 for streamlined GPU workload management
Enterprise Management - Base Command Manager 11.25.08 for cluster provisioning and workload orchestration

Complete list of infrastructure components with versions and documentation links:

Table 1 Supported Infrastructure Software#
Product	Description	Version
Core Infrastructure Drivers
NVIDIA GPU Data Center Driver	GPU driver with Blackwell (HGX B200, RTX PRO 6000), Hopper, Ada Lovelace, and Ampere architecture support	580.95.05
NVIDIA DOCA-OFED Driver for Networking	High-performance networking for InfiniBand and Ethernet	3.1.0
Virtualization
NVIDIA vGPU for Compute (Virtual GPU Manager and Guest Drivers)	Enterprise GPU virtualization with RTX PRO 6000 Blackwell Server Edition support on VMware vSphere	19.2
Container Platform
NVIDIA Container Toolkit	GPU-accelerated container runtime	1.18.0
Kubernetes Operators
NVIDIA GPU Operator	Automated GPU software lifecycle management for Kubernetes	25.10.0
NVIDIA Network Operator	Streamlined networking for GPU workloads in Kubernetes	25.7.0
NVIDIA DPU Operator (DPF)	Integrated DOCA Platform Framework for BlueField DPU lifecycle management	25.7.1
NVIDIA NIM Operator	Simplified deployment and scaling of NVIDIA Inference Microservices	3.0.1
Cluster Management & Orchestration
NVIDIA Base Command Manager (BCM)	Enterprise cluster provisioning and workload orchestration	11.25.08