AMD, Schneider Electric Develop AI Data Centre Reference Architecture

Fadekemi Ajakaiye

Schneider Electric and Advanced Micro Devices (AMD) have jointly developed a reference architecture for AI data centres, aiming to simplify the deployment of high-density computing infrastructure as demand for artificial intelligence workloads accelerates.

The companies said the engineering design provides data centre operators with a validated blueprint for deploying AMD’s Helios rack-scale platform, helping reduce the complexity and risks associated with building AI infrastructure.

The reference architecture marks the first outcome of the companies’ collaboration and is designed to support AI workloads running on AMD Helios, which integrates AMD Instinct MI455X graphics processing units, sixth-generation EPYC processors, Pensando Vulcano network interface cards and the ROCm software platform.

As AI models require greater computing power and energy, data centre operators face increasing challenges in managing electricity demand, cooling systems and facility design. Schneider Electric and AMD said prevalidated reference designs can shorten planning and deployment by providing tested guidance on integrating power, cooling and IT infrastructure.

“Today organisations require comprehensive, AI-ready reference designs that can take them from planning to deployment faster and with less risk,” said Manish Kumar, Executive Vice President of Secure Power and Data Centers at Schneider Electric.

According to the companies, the design supports modular AI clusters with up to 10.4 megawatts of IT capacity for new data centre developments and high-density computing environments consuming up to 246 kilowatts per rack.

The architecture also incorporates liquid-cooling technology from Motivair, a Schneider Electric company, using cooling distribution units and hybrid air-and-liquid systems capable of removing up to 84 per cent of generated heat.

To improve operational efficiency, the design integrates Schneider Electric’s ETAP and EcoStruxure IT Design simulation tools alongside AVEVA’s Unified Operations Center, enabling infrastructure modelling, real-time monitoring and predictive maintenance.

Forrest Norrod, Executive Vice President and General Manager of AMD’s Data Center Solutions Business Group, said AI infrastructure was evolving towards large-scale AI factories that require computing, networking, power and cooling systems to be designed together.

“By working with Schneider Electric to create a validated reference design, we are giving customers a practical blueprint to accelerate high-density AI deployments, reduce integration risk and scale with greater confidence and efficiency,” Norrod said.

The companies said the reference architecture has been validated to ANSI standards for deployments in the United States and will be expanded to support International Electrotechnical Commission standards for international markets.

The announcement comes as technology companies race to expand AI computing capacity, driving demand for more energy-efficient and scalable data centre infrastructure capable of supporting increasingly complex artificial intelligence applications.