AMD unveils Helion open rack blueprint for agentic AI infrastructure
- Advanced Micro Devices (AMD) introduced AMD Helion, an open, Open Compute Project-based rack blueprint designed to support flexible agentic AI infrastructure.
- The platform supports customizable CPU and GPU configurations, enabling customers to optimize performance, power consumption, and costs for different AI workloads.
- A CPU-only configuration powered by 6th Gen AMD EPYC processors can deliver up to 32,768 CPU cores for hyperscale agent execution.
- An all-in-one configuration combines AMD EPYC CPUs, AMD Instinct MI350P GPUs, and general-purpose computing for enterprise agentic AI workloads.
- AMD Helion supports both air-cooled and liquid-cooled designs and aims to integrate agentic AI capabilities into existing IT infrastructure.
- AMD plans to contribute the specification to the OCP community for further development and adoption.
- Helion rack configurations are under development with ecosystem partners and are planned to become available in November 2026.
- Major ecosystem partners supporting the initiative include Cisco (CSCO), Dell Technologies (DELL), Hon Hai, HPE (HPE), Lenovo, QCT, Supermicro, and Wiwynn.
KioskNews shows a cleaned-up reading view extracted from the publisher’s page — the original always lives on their site, not ours.