Red Hat AI Factory with NVIDIA

July 22, 2026•
Resource type: Datasheet
Print to PDF

Overview

As AI becomes instrumental to how organizations run and differentiate their business, they need greater control over applications, data, models, capacity planning, costs, and lifecycle management. Many teams are adopting AI through cloud services, disconnected tools, and fragmented infrastructure, making it harder to integrate AI into their own environments, apply proprietary data, manage usage, and govern AI consistently. 

Red Hat® AI Factory with NVIDIA is a co-engineered, supported, enterprise-grade AI solution that brings together Red Hat’s open hybrid cloud platform, AI engineering capabilities, and enterprise operations and NVIDIA’s optimized AI software, models, microservices, networking, and accelerated infrastructure. It helps organizations run AI with greater control, consistency, privacy, and predictability across hybrid cloud environments.

Features and benefits

Table 1. Features and benefits of Red Hat AI Factory with NVIDIA

Feature

Benefit

Co-engineered and jointly supported enterprise AI solution

Combines Red Hat AI Enterprise and NVIDIA AI Enterprise in an integrated, enterprise AI stack with production support from both Red Hat and NVIDIA.

Unified platform for enterprise AI

Provides an integrated stack that remains open across models, tools, infrastructure, and environments, helping reduce fragmentation without forcing a closed appliance model.

Optimized high-performance inference

Helps platform teams meet model-serving service level objectives (SLOs) with optimized capabilities for throughput, latency, density, and use of NVIDIA accelerated infrastructure.

Scalable distributed serving

Supports large and mixed-model inference workloads with distributed serving for enterprise users, applications, and rack-scale infrastructure where supported.

AI quickstart examples

  • Models-as-a-Service: Centrally host and serve approved models through API-based access.

  • Enterprise RAG: Ground AI applications in proprietary enterprise data.

  • Enterprise research: Use agentic workflows to search, synthesize, and distill internal content.

  • Video search and summarization: Search, analyze, and summarize video content at scale.

Enterprise data as an intelligence asset

Helps teams apply proprietary data through retrieval-augmented generation (RAG), customization, evaluation, synthetic data generation, and AI pipelines to create business-specific intelligence.

Validated blueprints and AI quickstarts

Provides co-engineered prototype samples and tested starting points that show how to combine Red Hat and NVIDIA capabilities for common AI use cases.

Agentic AI operations

Helps teams identify frameworks and deploy, observe, control, and govern agents while connecting them to enterprise systems, tools, data sources.

Security and governance for autonomous workflows

Supports guardrails, observability, auditability, isolation, and least-privilege access patterns for AI workloads and agents that can act on users’ behalf.

Predictable economics and infrastructure value

Supports planning with node-based pricing while helping organizations improve use, manage cost growth, and strengthen return on investment (ROI)/ total cost of ownership (TCO) arguments for AI infrastructure investments.

Technical specifications

Red Hat AI Factory with NVIDIA includes Red Hat AI Enterprise and NVIDIA AI Enterprise in a jointly supported offering for enterprise AI across hybrid cloud environments.

Key capabilities include:

  • Enterprise AI platform: Red Hat AI Enterprise delivers a tested and supported full AI stack, including Red Hat OpenShift®, to support model development, customization, evaluation, deployment, inference, observability, governance, and lifecycle management. It also provides capabilities for AI gateways, synthetic data generation, AI pipelines, metrics, tracing, logging, and alerting.
  • Inference and serving: Red Hat AI Inference, vLLM, llm-d, NVIDIA NIM microservices, NVIDIA Triton Inference Server, and NVIDIA TensorRT LLM support high-performance model serving and distributed inference where supported.
  • Models, agents, and customization: This includes Red Hat validated and optimized models, NVIDIA optimized open models, NVIDIA NeMo, enterprise RAG, and AgentOps capabilities that help teams customize models, connect models to enterprise data and tools, and manage governed agent lifecycles.
  • NVIDIA infrastructure enablement: NVIDIA CUDA-X libraries, NVIDIA GPU Operator, NVIDIA DOCA Operator, NVIDIA Network Operator, and NVIDIA NIM Operator help optimize and manage AI workloads on NVIDIA-accelerated computing and networking infrastructure.

Explore the foundation for enterprise AI

Discover how Red Hat AI Factory with NVIDIA helps organizations run AI with greater control, consistency, and performance across the hybrid cloud. Visit the Red Hat AI Factory with NVIDIA product page.

Start your AI Factory journey with trusted partners

Engage your preferred original equipment provider (OEM), solution provider, or distributor in the Red Hat and NVIDIA ecosystem to simplify procurement and map your unified AI platform strategy. Connect with a certified partner today.

Accelerate to production with Red Hat Services

Red Hat Consulting helps design initial AI architecture and onboard priority use cases, while Red Hat Training and Red Hat Technical Account Management help teams build GenAIOps skills and scale with confidence. Explore Red Hat Services for AI Factory.

blank png

Tags:AI inference, Artificial intelligence