NVIDIA BaseOS
Type: Platform Tags: NVIDIA, BaseOS, DGX OS, operating system, AI factory, Linux Related: NVIDIA-DGX, NVIDIA-DGX-Spark, NVIDIA-DGX-Station, NVIDIA-DGX-BasePOD, NVIDIA-DGX-SuperPOD, NVIDIA-Base-Command-Manager, NVIDIA-Data-Center-GPU-Drivers, NVIDIA-DCGM, CUDA-Compatibility Sources: https://docs.nvidia.com/baseos/index.html Last Updated: 2026-04-29
Summary
NVIDIA BaseOS defines production-ready operating system environments optimized for AI, machine learning, and data analytics workloads. It includes BaseOS software, DGX OS for Ubuntu-based DGX systems, and DGX EL for Red Hat Enterprise Linux environments. Current DGX product/docs coverage extends the OS context down to NVIDIA-DGX-Spark and NVIDIA-DGX-Station, where DGX OS or Ubuntu with NVIDIA AI Developer Tools provides the local AI development base.
Detail
Purpose
NVIDIA AI systems need validated operating-system layers with the right drivers, diagnostics, monitoring, and system configuration. BaseOS provides that foundation for DGX and AI factory deployments.
Key capabilities
- Optimized system configurations for AI workloads.
- Enhanced NVIDIA drivers, diagnostics, and monitoring tools.
- Support for Ubuntu, Red Hat Enterprise Linux, and Rocky Linux contexts.
- DGX OS as a customized Ubuntu distribution for DGX systems.
- DGX EL path for Red Hat Enterprise Linux environments.
- Ecosystem integration guides, including Red Hat OpenShift.
NVIDIA context
BaseOS sits under NVIDIA-DGX, NVIDIA-DGX-SuperPOD, NVIDIA-Base-Command-Manager, and NVIDIA-AI-Enterprise. It is a practical source of truth for the validated OS layer in NVIDIA AI infrastructure.
Connections
- NVIDIA-DGX - DGX systems rely on validated OS images and release notes.
- NVIDIA-DGX-Spark - compact DGX OS-based desktop system.
- NVIDIA-DGX-Station - Ubuntu with NVIDIA AI Developer Tools and DGX-style validated software.
- NVIDIA-DGX-BasePOD - BasePOD deployment guides use validated DGX OS/BaseOS and management stacks.
- NVIDIA-DGX-SuperPOD - cluster deployments need consistent OS foundations.
- NVIDIA-Data-Center-GPU-Drivers - driver packages are a core BaseOS concern.
- NVIDIA-DCGM - diagnostics and monitoring are part of system operations.
- CUDA-Compatibility - OS/driver/toolkit compatibility affects workload deployment.
Source Excerpts
- NVIDIA BaseOS docs describe production-ready operating system environments optimized for AI and data analytics workloads.