Introducing AMD Helion, Agentic AI Compute at Rack Scale

AMD introduced the AMD Helion agentic AI rack, an open blueprint for agentic AI infrastructure based on an Open Compute Project specification design and powered by AMD compute. The solution builds on the open-platform approach of the AMD Helios rack-scale solution, offering partners and customers flexibility for agentic AI deployments. AMD Helion allows customers to align CPU performance, core density and power with GPU acceleration tailored to the specific demands of their AI agent workloads.

AMD Helion is designed to support diverse agentic AI workloads, including personal agents, enterprise applications, code execution and complex workflows, each with distinct compute requirements. Dan McNamara, senior vice president and general manager of Compute and Enterprise AI at AMD, emphasized the need for optimized and adaptable infrastructure to meet varying workload demands. The solution integrates AMD EPYC performance, open software and flexible rack designs to enable partners to build systems aligned with customer needs while optimizing power usage.

The AMD Helion rack blueprint, based on an OCP specification, supports open software and offers flexible configurations, including liquid- and air-cooled versions. Customers can deploy agentic AI within existing IT infrastructure, repurposing prior investments to support new AI capabilities. Partners can customize CPU and GPU configurations to match specific workloads, balancing agent execution, model inference and supporting compute with performance, power and cost requirements.

AMD plans to contribute the AMD Helion specification to the Open Compute Project, making it available for community development and broader adoption. AMD Helion agentic rack configurations are currently in development with ecosystem partners and are expected to be available in November 2026. The initiative reflects AMD’s commitment to open, flexible AI infrastructure platforms and collaboration across the ecosystem.