Pranav Hotkar
San Francisco - July 23, 2026 - Schneider Electric and AMD have introduced the first reference design for the AMD Helios AI platform, providing data center developers with a standardized blueprint for deploying high-density AI infrastructure as demand for large-scale GPU clusters continues to accelerate.
The jointly developed architecture is designed to simplify the construction of AI factories by integrating compute, power distribution, cooling, and rack infrastructure into a validated deployment model. The companies said the reference design will enable operators to shorten deployment timelines while supporting the increasingly complex power and thermal requirements of next-generation AI workloads.
Built around AMD's upcoming Helios rack-scale platform, the reference design incorporates AMD Instinct MI455X GPUs, 6th Generation AMD EPYC processors, and AMD Pensando Vulcano network interface cards. On the infrastructure side, Schneider Electric contributes power distribution, electrical systems, cooling technologies, and digital management capabilities optimized for AI data center environments.
The design supports AI racks with power densities of up to 246 kW and modular AI clusters capable of scaling to approximately 10.4 MW of IT load, addressing one of the industry's biggest challenges as GPU-powered computing drives rack power requirements well beyond those of traditional enterprise data centers. Higher-density deployments have increased demand for integrated power and cooling systems capable of maintaining operational efficiency while reducing construction complexity.
Rather than serving as a finished product, the Helios reference architecture provides engineering guidance that data center developers, hyperscale cloud providers, and colocation operators can use to accelerate AI campus construction. By validating infrastructure components in advance, Schneider Electric and AMD aim to reduce design risks, improve interoperability, and speed deployment of AI-ready facilities.
The announcement reflects a broader shift across the digital infrastructure sector toward pre-engineered AI data center designs. As organizations invest billions of dollars in generative AI infrastructure, vendors are increasingly collaborating to deliver integrated platforms that combine advanced processors with standardized electrical and mechanical infrastructure, reducing time-to-market for new facilities.
For Schneider Electric, the collaboration strengthens its position in AI data center infrastructure by extending its portfolio of power management, cooling, and automation technologies. For AMD, the reference design provides customers with a deployment framework tailored to its next-generation AI computing platform, helping accelerate adoption of Helios-based systems.
With AI infrastructure projects growing in both scale and power density, standardized reference architectures are becoming an increasingly important tool for the industry. By aligning compute hardware with validated data center infrastructure, Schneider Electric and AMD are seeking to help operators deploy AI factories more efficiently while addressing the power, cooling, and scalability demands of next-generation high-performance computing environments.
.jpg)
