AMD showcases Helios AI rack platform at OCP Global Summit

AMD (NASDAQ: AMD) displayed its "Helios" rack-scale platform at the Open Compute Project Global Summit in San Jose. The platform is built on the Open Rack Wide specification introduced by Meta and integrates AMD Instinct GPUs, EPYC CPUs, and Pensando networking components.

The Helios platform adopts the new Open Rack Wide standard, which defines a double-wide rack designed for power, cooling, and serviceability requirements of AI systems in gigawatt-scale data centers. The system incorporates open compute standards including OCP DC-MHS, UALink, and Ultra Ethernet Consortium architectures.

"Open collaboration is key to scaling AI efficiently," said Forrest Norrod, executive vice president and general manager of AMD's Data Center Solutions Group. "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."

The rack features quick-disconnect liquid cooling for thermal management, a double-wide layout for improved serviceability, and standards-based Ethernet for multi-path resiliency. As a reference design, Helios enables original equipment manufacturers, original design manufacturers, and hyperscalers to customize open AI systems.

The platform supports both scale-up and scale-out fabric configurations and is designed to reduce deployment time while improving interoperability for AI and high-performance computing workloads. The system represents AMD's collaboration with the OCP community to develop open, scalable infrastructure for AI deployments.

You May Also Be Interested In

• Polar Power adds two directors with defense and business experience
• MongoDB (MDB) PT Raised to $519 at Citizens on Improved Q2 Sales, Positive Diligence Data
• Trump highlights ibogaine development push, Psyence BioMed responds