Global open-source solutions company Red Hat said on Wednesday it is expanding cooperation with Nvidia and will work to spread security-enabled, enterprise-grade agentic AI across open hybrid clouds through Red Hat AI Factory with NVIDIA.

The company said Red Hat is developing the OpenShell Reference Architecture with Nvidia to secure safety and control at the core of enterprise agentic AI deployments. Red Hat AI Factory with NVIDIA supports this.

The OpenShell Reference Architecture is based on OpenShell, an open-source project and agent sandbox runtime for which Red Hat and Nvidia are key contributors.

It combines Model-as-a-Service (MaaS) access control, an AI Gateway, and high-performance inference functions via vLLM and llm-d with NVIDIA Dynamo, NVIDIA NIM microservices and GPU orchestration technology. As the use of AI agents expands, demand for inference is also rising sharply, and the importance of optimized execution layers and efficient model access methods is increasing. With this in mind, Red Hat and Nvidia explained they developed a joint reference design based on Red Hat AI Factory with NVIDIA.

The two companies provide standardized guardrails, operational visibility and runtime safety using open-source community technologies.

Red Hat is strengthening hardware cooperation with Nvidia so enterprise infrastructure can respond to high-load agent workloads. It will support next-generation architectures, including NVIDIA Vera Rubin, from the day they are launched.

Red Hat Enterprise Linux for NVIDIA also supports the NVIDIA Vera CPU, providing performance optimized for agent processing and compute execution.

Red Hat is also working with Nvidia and the community to integrate the OpenShell Reference Architecture as a default platform capability into Red Hat AI and Red Hat AI Factory with NVIDIA.

Joe Fernandes (조 페르난데스), vice president and general manager of Red Hat's AI business unit, said, "Strengthening agentic AI security will lead the next wave of innovation in enterprise computing. But scaling autonomous agents requires trust, safety and seamless infrastructure integration." He said, "Red Hat is strengthening cooperation with Nvidia in open-source communities such as OpenShell, and is providing optimized inference architectures and day-one hardware support. Through this, we are providing a trusted foundation that enables IT decision-makers to deploy AI agents at scale with confidence."