NVIDIA AI Enterprise Infra 6.4 Release Notes#

Release Highlights#

Expanded Platform Support - Added support for Red Hat Enterprise Linux (RHEL) with KVM versions 9.6 and 10, and VMware vSphere (ESXi) version 9.0
GPU Infrastructure - NVIDIA GPU Data Center Driver 570.158.01 provides enhanced GPU acceleration and management capabilities
Kubernetes Orchestration - GPU Operator 25.3.1, Network Operator 25.4.0, and NIM Operator 2.0.0 deliver streamlined deployment and lifecycle management for GPU workloads
Networking Performance - DOCA-OFED Driver 25.4.0 (DOCA 3.0.0) for high-performance InfiniBand and Ethernet networking
Virtualization - vGPU for Compute 18.3 enables enterprise GPU virtualization across supported hypervisors
Platform Lifecycle - Discontinued support for RHEL with KVM versions 8.8, 9.2, and 9.5 as part of standard lifecycle management

Complete list of infrastructure components with versions and documentation links:

Table 4 Supported Infrastructure Software#
Product	Description	Version
Core Infrastructure Drivers
NVIDIA GPU Data Center Driver	GPU driver with Hopper, Ada Lovelace, and Ampere architecture support	570.158.01
NVIDIA DOCA-OFED Driver for Networking	High-performance networking for InfiniBand and Ethernet	3.0.0
Virtualization
NVIDIA vGPU for Compute (Virtual GPU Manager and Guest Drivers)	Enterprise GPU virtualization with enhanced performance and management capabilities	18.3
Container Platform
NVIDIA Container Toolkit	GPU-accelerated container runtime with enhanced security	1.17.8
Kubernetes Operators
NVIDIA GPU Operator	Automated GPU software lifecycle management for Kubernetes	25.3.1
NVIDIA Network Operator	Streamlined networking for GPU workloads in Kubernetes	25.4.0
NVIDIA NIM Operator	Simplified deployment and scaling of NVIDIA Inference Microservices	2.0.0
Cluster Management & Orchestration
NVIDIA Base Command Manager (BCM)	Enterprise cluster provisioning, workload orchestration, and lifecycle management	11.25.05