Schneider Electric and AMD have jointly developed a reference design for the Helios rackscale solution aimed at targeting the deployment of high-density AI data centre environments.
The new blueprint supports AI racks with power densities reaching up to 246kW, along with modular AI clusters designed to handle IT loads of up to 10.4MW.
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The design is specifically made for the AMD Helios rackscale solution, which incorporates AMD Instinct MI455X graphics processing units (GPUs), 6th Gen AMD EPYC central processing units (CPUs), and AMD Pensando Vulcano network interface cards.
According to Schneider Electric, the reference architecture addresses facility power, facility cooling, IT space, and lifecycle software technical domains.
Liquid cooling is a core element of the design, using Motivair by Schneider Electric cold distribution unit-based and hybrid air/liquid solutions that reportedly remove up to 84% of heat.
Electrical and thermal validation of the architecture was carried out using electrical transient and analysis programme (ETAP) and EcoStruxure IT Design computational fluid dynamics (CFD) simulation tools.
These include integrated Electrical Digital Twin capabilities for infrastructure performance management.
The companies state that power usage effectiveness (PUE) can be as low as approximately 1.12 at full load.
The joint development aims to provide a tested and scalable template for data centre architects and operators tackling high-density AI workloads.
Pre-validated reference designs like this are put forward as a means to reduce risk and complexity in building reliable and scalable data centre environments capable of accommodating the latest AI systems.
Schneider Electric secure power and data centres executive vice-president Manish Kumar said: “Today organisations require comprehensive, AI-ready reference designs that can take them from planning to deployment faster and with less risk.
“Through our collaboration with AMD, we’re delivering an engineering-backed reference design that bridges the gap between advanced AI compute platforms, energy tech, and real-world data centre implementation, enabling customers to deploy scalable, high-density AI environments with greater confidence, efficiency, and speed.”
The design has been validated to American National Standards Institute (ANSI) standards for US deployment, with intentions to adapt the framework to International Electrotechnical Commission (IEC) standards for international use in the future.
In February 2026, AMD and Tata Consultancy Services (TCS) announced plans to jointly develop a new rack-scale AI infrastructure in India, using the AMD “Helios” platform.
