Modular Liquid Cooling and Power Blueprint Developed for Hyperscale AI Data Centers

Siemens and nVent collaborate on a Tier III-capable reference architecture that combines industrial electrical systems with advanced liquid cooling to accelerate AI infrastructure deployment and efficiency. By Embedded Systems Engineering / 15 Dec 2025
Follow ESE

Siemens and nVent have announced a joint reference architecture for liquid cooling and power infrastructure, designed specifically for hyperscale AI workloads. The blueprint aims to support data center operators in deploying AI infrastructure faster, more sustainably, and with greater intelligence.

The new architecture is engineered to help customers build 100 MW hyperscale AI data centers intended to house large-scale, liquid-cooled AI infrastructure, such as the NVIDIA GB200 NVL72 systems. The Tier III-capable design integrates Siemens’ industrial-grade electrical and automation systems with NVIDIA DGX GB200 reference designs and liquid cooling technology from nVent.

As AI workloads become more intensive and distributed, data center strategies must prioritize performance, efficiency, and scalability through increasingly adaptive systems. The collaboration is intended to help data centers prepare their cooling and power infrastructure for global deployment and operational resilience.

Ciaran Flanagan, Global Head of Data Center Solutions at Siemens, stated that the reference architecture “accelerates time-to-compute and maximizes tokens-per-watt, which is the measure of AI output per unit of energy.” He described the blueprint as being “modular, fault-tolerant, and energy-efficient,” and emphasized that the joint effort is connecting the dots across the value chain to drive innovation and interoperability, helping operators build future-ready data centers.

“We have decades of expertise supporting customers’ next-generation computing infrastructure needs,” said Sara Zawoyski, president of nVent Systems Protection. She added that the collaboration with Siemens underscores that commitment, noting, “The joint reference architecture will help data center managers deploy our cutting-edge cooling infrastructure to support the AI buildout.”

Data centers currently face challenges including rising rack-level power densities, more compute-intensive workloads, and the need for modularity to maintain uptime and scalability. Reference architectures are seen as critical tools for aiding operators in rapid deployment and interface standardization while providing a framework for innovation among infrastructure providers.

Siemens contributes expertise in industrial-grade electrical systems and intelligent infrastructure, offering a portfolio that includes medium and low voltage power distribution, advanced automation, and energy management software. This allows for reliable and efficient operation of mission-critical facilities. nVent is providing its liquid cooling technology, leveraging its experience in solving complex cooling challenges for global cloud service providers and supporting high-density computing needs.

Posted by Embedded Systems Engineering Connect

Latest Articles

Hybrid Controller Simplifies USB-C Power Delivery Integration

STMicroelectronics introduces a patented hybrid mode to streamline software development for advanced USB-C sink applications.

Jan 27, 2026
ADLINK Integrates Intel Core Ultra Series 3 into New COM Express Modules for Edge AI

New hardware architecture delivers up to 180 platform TOPS and extended temperature support for mission-critical robotics and industrial automation.

Jan 23, 2026
Microchip Expands PolarFire FPGA Ecosystem with Low-Power SDI and CoaXPress Video Solutions

New IP cores and hardware bridging kits aim to streamline high-bandwidth video pipelines for medical, industrial, and robotic vision applications.

Jan 22, 2026
congatec Integrates AMD Ryzen AI P100 Processors into New Rugged COM Express Modules

A new series of Type 6 Compact modules delivers up to 59 TOPS of AI performance for mission-critical edge applications in harsh industrial environments.

Jan 20, 2026
ThunderSoft Debuts AI-Native Operating System Architecture Across Automotive and Robotics Sectors

New AIOS platform integrates system-level artificial intelligence to bridge the gap between digital models and physical hardware execution.

Jan 12, 2026
Modular Development Kit Streamlines Zone Controller Design for Software-Defined Vehicles

Infineon and Flex debut a scalable ZCU platform featuring 30 unique building blocks to accelerate automotive E/E architecture transitions.

Jan 07, 2026

Featured Content

Modular Development Kit Streamlines Zone Controller Design for Software-Defined Vehicles

Infineon and Flex debut a scalable ZCU platform featuring 30 unique building blocks to accelerate automotive E/E architecture transitions.

Jan 07, 2026
NXP Launches S32N7 Processor Series to Centralize Core Vehicle Functions

New 5 nm super-integration processors enable OEMs to consolidate propulsion, safety, and body domains into a single hub, reducing architecture complexity and costs.

Jan 05, 2026
High-Performance Edge Platform Debuts to Accelerate AI-Driven Clinical Diagnostics

Axiomtek introduces a medical-grade computing system featuring 13th Gen Intel Core processing and dedicated GPU expansion for real-time imaging and surgical monitoring.

Jan 04, 2026