AMD introduces open Helion rack blueprint for agentic AI

Key Takeaways

• AMD Helion is an open‑source rack design built on the Open Compute Project (OCP) spec, targeting agentic AI workloads.
• The blueprint supports both air‑cooled and liquid‑cooled builds and can be tailored with up to 32,768 CPU cores or a hybrid CPU + GPU configuration.
• Two reference models are defined: a CPU‑only hyperscale rack (6th‑Gen EPYC) and an all‑in‑one rack that adds Instinct MI350P GPUs.
• AMD will submit the Helion spec to OCP in Q4 2026, enabling ecosystem partners to customize and mass‑produce the solution.

Introduction: An Open Blueprint for Agentic AI

AMD has announced Helion, a rack‑level reference architecture that embraces the Open Compute Project’s design principles while delivering the compute density required for autonomous AI agents. By decoupling hardware selection from software stacks, Helion lets data‑center operators retrofit existing infrastructure and scale AI capabilities without vendor lock‑in.

Design Philosophy

Open‑Compute Foundations

• OCP‑compliant chassis – standard dimensions, power distribution, and management interfaces.
• Modular cooling – interchangeable liquid‑ or air‑cooling modules to match thermal budgets from 150 kW to 300 kW per rack.

Flexible Compute Mix

Helion’s blueprint lets customers balance CPU core count, GPU acceleration, and power envelope to match the specific demands of agent execution, model inference, and supporting workloads.

Reference Configurations

Numbers reflect the maximum supported density per the Helion spec; actual deployments may be sized lower to meet specific power or cooling constraints.

Deployment Benefits

1. Leverage Existing Assets – Operators can install Helion modules into current OCP racks, converting legacy power and networking investments into AI‑ready platforms.
2. Performance‑Cost Matching – Partners can select a CPU‑only or hybrid mix, aligning silicon spend with the workload’s compute‑to‑memory ratio.
3. Scalable Power Envelope – The design accommodates up to 300 kW per rack, with optional liquid‑cool loops that keep GPU temperatures below 70 °C under sustained AI loads.
4. Software Agnostic – Open‑source drivers, container runtimes, and orchestration tools (Kubernetes, OpenShift) are fully supported, simplifying integration with existing MLOps pipelines.

Ecosystem Collaboration

AMD is positioning Helion as a community‑driven standard. By contributing the specification to the OCP foundation, AMD invites OEMs, system integrators, and cloud providers to iterate on chassis designs, cooling solutions, and firmware. Early partners are already prototyping custom power distribution units (PDUs) that reduce idle power draw by 12 % compared with legacy rack designs.

Timeline

• Q3 2026 – Draft Helion spec released to OCP mailing list for public comment.
• Q4 2026 – Formal submission to OCP; reference hardware samples shipped to select partners.
• Q1 2027 – First commercial Helion racks expected in hyperscale and enterprise data centers.

Bottom Line

AMD’s Helion rack blueprint offers a transparent, OCP‑aligned pathway for deploying agentic AI at scale. By providing two well‑defined reference builds—one focused on massive CPU core density and another that couples EPYC processors with Instinct MI350P GPUs—Helion gives data‑center operators the levers needed to align performance, power, and cost with their AI strategies. With the specification slated for OCP release in late 2026, the industry can anticipate broader adoption, faster innovation cycles, and a more open AI infrastructure ecosystem.