Today at the Open Compute Project (OCP) Global Summit in San Jose, AMD (NASDAQ: AMD) showcased a static display of its “Helios” rack scale platform for the first time.
Developed based on the new Open Rack Wide (ORW) specification, introduced by Meta, AMD said Helios extends the AMD open hardware philosophy from silicon to system to rack, representing a major step forward in open, interoperable AI infrastructure.
Helios, first discussed by AMD in a blog last June, integrates open compute standards including OCP DC-MHS, UALink, and Ultra Ethernet Consortium (UEC) architectures, supporting open scale-up and scale-out fabrics. The rack includes quick-disconnect liquid cooling, a double-wide layout for serviceability, and standards-based Ethernet.
The company said Helio provides a foundation for open, scalable AI infrastructure. Designed to meet the demands of gigawatt-scale data centers, the new ORW specification is an open, double-wide rack for the power, cooling, and serviceability needs of AI systems. By adopting ORW and OCP standards, Helios provides a unified, standards-based foundation to develop and deploy AI infrastructure at scale, AMD said.
“Open collaboration is key to scaling AI efficiently,” said Forrest Norrod, executive vice president and general manager, Data Center Solutions Group, AMD. “With Helios, we’re turning open standards into real, deployable systems — combining AMD Instinct GPUs, EPYC CPUs, and open fabrics to give the industry a flexible, high-performance platform built for the next generation of AI workloads.”
As a reference design, Helios enables OEMs, ODMs, and hyperscalers to adopt, extend, and customize open AI systems quickly — reducing deployment time, improving interoperability, and supporting efficient scaling for AI and HPC workloads. The Helios platform reflects the ongoing collaboration from AMD across the OCP community to enable open, scalable infrastructure for AI deployments worldwide.
