Glossary

Key terms used throughout the NVIDIA AI Enterprise Software documentation.

Application Layer

The set of NVIDIA AI Enterprise software components for building and deploying AI solutions - including AI frameworks, SDKs, NVIDIA NIMs, NVIDIA Omniverse, and pre-trained models.

Feature Branch (FB)

A release branch containing the latest software versions. Designed for development, testing, and research environments with a rapid release cadence and short support window.

Government Ready Containers

Hardened container images that meet software security requirements for FedRAMP High or equivalent Sovereign use cases. Available on Production Branch (PB) and Long-Term Support Branch (LTSB) only - not on Feature Branch (FB) or Infrastructure Branch.

Infrastructure Branch

A release branch for GPU drivers and infrastructure operators. Major and minor releases follow a regular cadence, with standard and long-term support options available.

Infrastructure Layer

The set of NVIDIA AI Enterprise software components for managing GPU compute resources - including GPU drivers, NVIDIA GPU Operator, NVIDIA Network Operator, NVIDIA Container Toolkit, NVIDIA Run:ai, and cluster management tools.

Long-Term Support Branch (LTSB)

A release branch designed for highly regulated industries with extended certification cycles and strict change control requirements. Provides the longest support window with periodic security patches.

NVIDIA NGC Catalog

NVIDIA’s hub for GPU-optimized software. All NVIDIA AI Enterprise supported components are listed in the NGC Catalog and identified by the “NVIDIA AI Enterprise Supported” designation.

NVIDIA Omniverse

A platform for 3D collaboration, simulation, OpenUSD workflows, and digital twins. Included in NVIDIA AI Enterprise licenses as an Application Layer component.

NVIDIA Run:ai

A GPU orchestration and workload scheduling platform for Kubernetes environments, composed of the Run:ai Control Plane service (centralized management, UI, and policy) and Run:ai Cluster services. Both NVIDIA Run:ai self-hosted and Run:ai SaaS are included in NVIDIA AI Enterprise licenses, giving customers a choice between deploying and managing the control plane themselves or consuming an NVIDIA-managed Control Plane cloud service.

Production Branch (PB)

A release branch optimized for production deployments. Provides API stability, regular security patches, and a longer support window than Feature Branches.

Run:ai SaaS

NVIDIA-managed cloud deployment of NVIDIA Run:ai. Included in the NVIDIA AI Enterprise license under the same enterprise SLA as self-hosted NVIDIA Run:ai.

NVIDIA Base Command Manager

Cluster management and provisioning software for NVIDIA DGX systems and AI infrastructure at scale.

NVIDIA Container Toolkit

A set of runtime components and utilities that enable GPU-accelerated containers across container engines such as Docker, containerd, and CRI-O.

NVIDIA DOCA Microservices

A set of prebuilt, composable microservices that run natively on NVIDIA BlueField to accelerate AI factories with high-performance networking, cybersecurity, data acceleration, and orchestration.

NVIDIA DPU Operator (DPF)

A Kubernetes operator and framework that automates the provisioning, orchestration, and lifecycle management of NVIDIA BlueField DPUs and DOCA Microservices, enabling optimized North-South networking in Kubernetes environments.

NVIDIA GPU Operator

A Kubernetes operator that automates the management of all NVIDIA software components needed to provision and manage GPUs in Kubernetes clusters.

NVIDIA NeMo

An end-to-end platform for building, customizing, and deploying generative AI models including large language models, multimodal, speech AI, and vision.

NVIDIA Network Operator

A Kubernetes operator that automates the provisioning and management of NVIDIA ConnectX NICs and SuperNICs, optimizing high-speed East-West networking and RDMA-based data transfers within Kubernetes clusters.

NVIDIA NIM

Optimized microservices for accelerating foundation model deployment. Available on the NVIDIA NGC Catalog with enterprise support.

NVIDIA NIM Operator

A Kubernetes operator that enables cluster administrators to deploy and manage NVIDIA NIM microservices in Kubernetes.

NVIDIA Triton Inference Server

A multi-framework inference serving platform that supports multiple model formats, dynamic batching, and optimized backends for deploying AI models at scale.

Support Matrix

A reference that lists supported hardware, operating systems, hypervisors, and orchestration platforms for each NVIDIA AI Enterprise infrastructure release.