AMD Unveils Its Helion Agentic AI Rack To Provide Partners And Customers With Flexibility For Agentic AI Infrastructure

AMD Helion’s OCP spec-based rack blueprint supports open software and can come in flexible configurations, including liquid- and air-cooled versions, allowing customers to create virtually any configuration needed for their needs. Some examples of this include:

• A CPU-only rack powered by 6th Gen AMD EPYC™ processors that can support up to 32,768 CPU cores. This rack is designed for hyperscale customers, providing dedicated capacity for agent execution and connections to separate model inference infrastructure.
• An all-in-one rack that combines AMD EPYC CPU nodes, AMD Instinct™ MI350P GPU nodes and general-purpose compute on one platform. This rack is designed to provide partners and customers an all-in-one concept for enterprise agentic AI.

AMD Helion enables customers to deploy agentic AI in the IT infrastructure they already have, turning existing investments into a foundation for new AI capabilities. Partners can tailor CPU and GPU configurations to customer workloads, matching agent execution, model inference and supporting compute to each deployment’s performance, power and cost requirements.

AMD plans to contribute the AMD Helion specification to OCP, making it available to the community for further development and adoption. AMD Helion agentic rack configurations are currently under development with ecosystem partners and are planned to be available in November 2026.

Ecosystem Support

AMD is helping shape the next phase of AI infrastructure through open, flexible platforms, a broad portfolio of workload-optimized compute and collaboration across the ecosystem. AMD Helion brings that approach to agentic AI, helping partners match software and compute to the work agents perform and giving customers flexibility to scale as their requirements evolve.