Schneider Electric and AMD have introduced the first validated reference design for the AMD Helios rack-scale platform, providing data centre operators with a ready blueprint to deploy high-density artificial intelligence (AI) infrastructure more quickly and with lower implementation risk. The jointly developed solution is designed to simplify the rollout of AI factories by combining advanced computing technology with integrated power, cooling and digital infrastructure.
The reference design is the first outcome of the strategic collaboration between the two companies and is intended to help enterprises and hyperscale operators accelerate the deployment of large-scale AI environments while improving energy efficiency, scalability and operational reliability.Built for the AMD Helios rack-scale platform, the reference design supports AI workloads powered by AMD Instinct MI455X GPUs, 6th Generation AMD EPYC processors, AMD Pensando Vulcano network interface cards, and the open ROCm software ecosystem. The platform has been engineered to deliver high-performance computing capabilities required for next-generation AI applications while optimising power consumption and system efficiency.
The solution supports AI racks with power densities of up to 246 kilowatts and enables modular AI clusters with information technology loads of up to 10.4 megawatts, allowing organisations to expand capacity as computing requirements grow. The design can be deployed in both newly built AI facilities and existing data centres undergoing high-density infrastructure upgrades.
As demand for AI computing continues to increase, data centres are facing new challenges related to power distribution, cooling efficiency and infrastructure planning. The validated reference design addresses these requirements by providing pre-engineered guidance covering facility power systems, cooling infrastructure, IT space planning and lifecycle management software. By reducing engineering complexity during the planning stage, organisations can shorten deployment timelines and improve operational performance.
Commenting on the collaboration, Manish Kumar, Executive Vice President – Secure Power & Data Centers, Schneider Electric, said organisations require practical, AI-ready infrastructure designs that simplify deployment while ensuring reliability and scalability. He noted that the collaboration with AMD combines advanced AI computing with Schneider Electric’s expertise in power management, cooling technologies and digital infrastructure, enabling customers to deploy high-density AI environments with greater confidence and efficiency.
Forrest Norrod, Executive Vice President and General Manager, Data Center Solutions Business Group at AMD, said AI infrastructure is rapidly evolving into large-scale AI factories where computing, networking, power and cooling must be designed as an integrated system. He added that the validated reference design provides customers with a practical framework to accelerate AI deployments while reducing integration challenges and improving operational efficiency.
The reference architecture incorporates advanced liquid cooling technologies from Motivair by Schneider Electric, capable of removing up to 84 per cent of generated heat through a combination of liquid and hybrid cooling systems. It also integrates Schneider Electric’s ETAP and EcoStruxure IT Design software for electrical and thermal modelling, digital twin capabilities and predictive infrastructure management, while AVEVA’s Unified Operations Center provides real-time operational monitoring and system visibility.
The reference design has been validated to ANSI standards for deployments in the United States, with future plans to extend support for IEC standards to enable wider adoption across global markets. The collaboration marks an important step towards delivering efficient, scalable and sustainable AI infrastructure for the rapidly growing data centre industry.
