Time:2026-07-29 Browse: 0
New AI infrastructure reference design provides a validated blueprint for high-density AI clusters, supporting scalable deployment with optimized power, cooling, and digital infrastructure solutions
July 2026 — Schneider Electric and AMD have officially introduced their jointly developed and validated AMD Helios rack-scale AI platform reference design, marking the first major milestone in their strategic collaboration to accelerate the deployment of next-generation AI factories.
Designed for high-density artificial intelligence workloads, the new reference design provides data center operators and enterprises with a proven architecture model to reduce deployment complexity, lower integration risks, and accelerate the delivery of AI-ready infrastructure.
The solution combines AMD’s advanced AI computing platform with Schneider Electric’s expertise in power distribution, thermal management, and digital infrastructure, creating an integrated approach for building scalable AI environments.
The newly released reference design is based on the AMD Helios rack-scale AI platform and is designed to support demanding AI workloads requiring massive computing capacity, high-speed interconnects, and advanced energy management.
The modular AI cluster architecture supports:
Up to 246 kW AI rack power density
Maximum 10.4 MW IT load capacity
Flexible expansion for large-scale AI factory deployments
Optimized power, cooling, and IT infrastructure integration
The platform supports AMD Instinct™ MI455X GPUs, sixth-generation AMD EPYC™ CPUs, AMD Pensando™ Vulcano SmartNICs, and the open ROCm™ software ecosystem.
By improving computing performance, memory capacity, networking bandwidth, and system-level integration, AMD Helios enables enterprises to run larger and more complex AI workloads while improving energy efficiency.

As AI applications continue to increase infrastructure requirements, traditional data center designs face new challenges related to power density, cooling capacity, and operational complexity.
The Schneider Electric and AMD reference design provides a validated framework that helps data center architects and operators plan AI infrastructure more efficiently.
Through advanced modeling and simulation of physical infrastructure performance, the reference design helps standardize:
Electrical power distribution architecture
Advanced cooling systems
AI-ready IT room design
Digital infrastructure management
This approach reduces engineering cycles and helps organizations build highly reliable, scalable, and AI-ready data centers.
The AMD Helios reference design covers four critical infrastructure areas:
Schneider Electric provides advanced electrical infrastructure solutions designed to support the extreme power demands of high-density AI computing environments.
The design integrates digital power management technologies to improve reliability, monitoring capabilities, and operational efficiency.
To manage increasing AI rack temperatures, the reference design incorporates advanced liquid cooling technologies from Motivair, a Schneider Electric company.
The cooling architecture uses a coolant distribution unit (CDU) design combined with hybrid air-liquid thermal management solutions, capable of removing up to 84% of generated heat.
The solution supports both new AI factory construction and upgrades of existing high-density computing facilities.
Through validated power and cooling configurations, enterprises can simplify deployment while reducing engineering uncertainty.
The reference design integrates Schneider Electric’s digital tools, including:
ETAP and EcoStruxure™ IT Design CFD simulation tools for power and thermal design validation
Digital twin capabilities for infrastructure modeling and performance analysis
AI-driven predictive maintenance and optimization
AVEVA unified operations center for real-time monitoring and visualization
These technologies enable operators to manage AI infrastructure throughout its complete lifecycle.
Energy efficiency has become a critical factor in large-scale AI deployment. According to Schneider Electric, the AMD Helios reference design can achieve a power usage effectiveness (PUE) level as low as approximately 1.12 under full-load operation, helping reduce energy consumption while maintaining high computing performance.
The solution has completed validation based on American National Standards Institute (ANSI) standards, with future plans to expand the framework to support IEC standards for global deployment.
Manish Kumar, Executive Vice President of Schneider Electric and Head of Secure Power and Data Center Business, stated that enterprises require AI-ready reference designs to reduce risks and shorten deployment timelines.
Through strategic cooperation with AMD, Schneider Electric aims to bridge the gap between advanced AI computing platforms, energy technologies, and real-world data center implementation.
Forrest Norrod, Executive Vice President and General Manager of AMD Data Center Solutions Business Unit, noted that AI infrastructure is rapidly evolving toward large-scale AI factories, where computing power, networking, power systems, and cooling must be designed together from the beginning.
The cooperation between Schneider Electric and AMD provides customers with a practical reference architecture for deploying high-density AI environments with greater efficiency, reliability, and scalability.
The launch of the AMD Helios AI platform reference design highlights the increasing importance of integrated solutions in the AI infrastructure market.
As enterprises accelerate AI adoption, successful deployment will depend not only on computing performance but also on efficient power management, advanced cooling technologies, intelligent monitoring, and sustainable data center operations.
Through the combination of AMD’s AI computing innovation and Schneider Electric’s energy management expertise, the new reference design provides a foundation for the next generation of AI factories worldwide.
Copyright © 2018-2025 Qunlebu Co., Ltd. All Rights Reserved. Excellent PLC GLB PLC MTS PLC