Infrastructure ·
AMD unveils Helion, a server rack designed for agentic AI
At the OCP Global Summit in San Jose, AMD showed Helion, a reference rack design for running AI agents. The CPU-only version reaches up to 32,768 cores.
Helion uses 6th-generation EPYC processors and can be combined with Instinct GPUs, so customers can balance agent execution, model inference and cost. AMD plans to contribute the specification to the Open Compute Project.
First configurations are expected in November 2026 from partners such as Dell, HPE, Lenovo and Supermicro. Power figures and pricing have not been published.
What it means for companies
Agents need more than large models: tool calls and data work take a lot of ordinary compute. For companies that want to keep data in-house, this opens a path to running agents on their own infrastructure.
Sources: DataCenterDynamics, StorageReview