Red Hat and NVIDIA unveil an open source–driven AI Factory that turns hybrid infrastructure into production-ready AI at scale, cutting complexity, cost and lock-in for enterprises.
Red Hat has extended its enterprise open source strategy into large-scale artificial intelligence with the launch of Red Hat AI Factory with NVIDIA, a co-engineered software platform designed to move organisations from AI pilots to full production.
The offering combines Red Hat AI Enterprise and NVIDIA AI Enterprise into an end-to-end stack for inference, tuning, deployment and operations, positioning open source software as the control layer for emerging “AI factories”. It targets agentic workloads, high-density inference and the growing strain on infrastructure as enterprise AI spending races toward a projected $1 trillion by 2029.
Built on Red Hat Enterprise Linux, open models and an open inference ecosystem including vLLM, the platform supports on-premises, cloud and edge deployments while reducing vendor lock-in. Pre-configured models such as IBM Granite, NVIDIA Nemotron and NVIDIA Cosmos are delivered as NVIDIA NIM microservices, with NVIDIA NeMo for faster tuning.
Operational features include unified high-performance serving, observability, GPU pooling, intelligent orchestration, automatic checkpointing and a security-hardened, zero-trust architecture powered by NVIDIA DOCA, aimed at lowering total cost of ownership.
Chris Wright, chief technology officer and senior vice president, Global Engineering, Red Hat, said: “The shift from AI experimentation to industrial-scale, enterprise-wide production requires a fundamental change in how we manage the AI computing stack. We’re accelerating the path to deploy AI and move quickly to production using Red Hat AI Factory with NVIDIA.”
The platform is validated across infrastructure from Cisco, Dell Technologies, Lenovo and Supermicro, and is available now.
