Data Centre Liquid Cooling for the AI Era
Advanced liquid cooling for high-density data centres. Efficient, scalable, and built for AI workloads.
Schneider Electric's data centre liquid cooling solutions are purpose‑built for AI workloads, GPU servers, and high‑density IT environments. With over a decade of experience cooling racks beyond 400 kW, we deliver end‑to‑end liquid cooling, with advanced technologies like Coolant Distribution Units (CDUs) and direct‑to‑chip systems, ensuring fast, scalable, and resilient AI data centre deployments.
In the news

In the news

Discover how liquid cooling works, how to adopt it, and how it powers AI and high-density IT. Access expert guides, white papers, and blogs.
Optimise your data centre with advanced design and simulation tools, along with expert consulting. Learn how to boost efficiency, reduce operational costs, and improve reliability with ours end‑to‑end cooling solutions.

Data centre reference designs
Revolutionise your infrastructure with validated data centre reference designs engineered for high‑density AI workloads.
Digital twins and simulations
Use digital twins for data centre management to simulate thermal behaviour, optimise liquid cooling performance, and reduce deployment risks.
EcoConsult
Our specialists analyse and map your assets and systems to boost efficiency and sustainability, using software‑driven insights to optimise energy use, enhance liquid cooling performance, and extend asset lifespan.Revolutionise your data centre infrastructure with cutting-edge AI-specific data centre solutions and meet the demands of high-compute workloads.

ChilledDoor® Rack Cooling System delivers 75kW rack-level liquid cooling with exceptional heat removal. Designed for AI and HPC, it integrates air and liquid cooling for sustainable thermal management and seamless Open19/OCP compatibility.
Our HDU enables liquid-cooled HPC systems to reject heat to air when chilled water isn't available. With EC fans, PLC control, and no building water dependency, it ensures efficient, flexible thermal management for high-performance data centres.
Optimise AI and HPC performance with Motivair by Schneider Electric's CDUs—advanced liquid cooling solutions engineered for reliability, high-density workloads, and sustainable thermal management.
Our advanced liquid cooling ensures peak thermal performance for AI and high-density workloads, keeping your data centre efficient even under extreme compute demands.
Explore our liquid and air cooling solutions
Use real-time analytics to detect inefficiencies, maximise uptime, and keep your AI infrastructure running at peak performance even under high-density workloads.

EcoStruxure™ Building Operations
Increase energy and operational efficiency with seamless monitoring, management, and control of building systems from one convenient, mobile-enabled control centre.
Data Centre Expert
A scalable monitoring software that collects, organises, and distributes critical device information providing a comprehensive view of equipment._BUILDER.IORectangle.webp)
IT Design CFD
EcoStruxure IT Design CFD empowers you to design highly efficient, optimally cooled data centres.Unlock end-to-end value—from infrastructure to intelligence—with trusted EcoXpert™ partners delivering scalable, future-ready AI Data Centre solutions tailored to your needs.
Need help?
Product Selector
Quickly and easily find the right products and accessories for your applications.
Get a Quote
Start your sales inquiry online and an expert will connect with you.
Where to buy?
Easily find the nearest Schneider Electric distributor in your location.
Help Centre
Find support resources for all your needs, in one place.
What is liquid cooling for data centres?
Liquid cooling is a high‑efficiency method that uses liquid (rather than air) to remove heat from servers, AI accelerators, and high‑density IT equipment. Because liquid captures heat far more effectively, it supports AI data centre cooling, higher rack densities, improved energy efficiency, and more sustainable high‑performance compute environments.
Related insights:
Read why liquid cooling is becoming essential for high-density AI data centres
Why is liquid cooling important for AI and GPU-based data centres?
AI, GPU, and accelerated computing clusters generate extreme heat that traditional air cooling cannot manage at scale. Liquid cooling delivers precise, chip‑level heat removal, preventing GPU throttling and maintaining predictable performance across high‑density AI racks and next‑generation compute architectures.
Related insights:
What are the main liquid cooling techniques used in AI data centres?
Data centres are increasingly turning to liquid cooling to support AI, HPC, and high‑density compute. Two main techniques dominate:
-Direct-to-Chip (DTC) liquid cooling, also called direct liquid cooling (DLC). This method uses cold plates to circulate coolant directly over CPUs, GPUs, and accelerators. It delivers high thermal efficiency and is now the leading solution for both brownfield and greenfield data centre deployments.
-Immersion Cooling. Servers are submerged in a dielectric fluid, enabling uniform heat removal with minimal reliance on airflow. It is used selectively where its unique performance or operational benefits outweigh design trade-offs.
Single-phase DTC cooling accounts for the majority of new liquid-cooled capacity, supported by advances in cold plate design as chip heat loads rise. Two-phase DTC adoption is expected to grow once chip TDPs exceed single-phase limits, but for now remains focussed on pilots. Immersion cooling continues to expand in targeted, high‑density environments.
Related insights:
-The rise of direct-to-chip cooling as a top AI cooling system
-Rethinking data centre cooling for AI with direct-to-chip liquid cooling
At what rack density does air cooling become ineffective for AI workloads?
Air cooling becomes insufficient as rack densities exceed 80–100 kW, which is common for modern GPU‑intensive and AI training environments. At these power levels, airflow cannot extract heat efficiently, making liquid‑based AI cooling systems the preferred choice for stable, high‑performance compute.
Related insights:
How to optimise air cooling for NVIDIA AI servers
Can existing air-cooled data centres be retrofitted with liquid cooling?
Yes. Most air‑cooled facilities can be upgraded with liquid cooling solutions such as direct‑to‑chip loops, rear‑door heat exchangers (RDHx), Heat Dissipation Units (HDUs), Coolant Distribution Units (CDUs), and hybrid AI cooling architectures, enabling support for high‑density AI racks without requiring a full data centre rebuild.
Related insights:
-Why liquid cooling reference designs matter at high densities
-Upgrade legacy data centres for AI workloads with RDHx liquid cooling
How does liquid cooling improve efficiency and sustainability in AI data centres?
Direct-to-chip liquid cooling reduces energy use by 30% to 60%, increases heat‑transfer performance, and lowers reliance on mechanical cooling systems. This improves PUE, reduces operational emissions, and supports sustainable AI infrastructure, aligning with long‑term decarbonisation and energy‑efficiency targets.
Related insights:
How CDUs improve data centre energy efficiency
Can liquid cooling scale to support next-generation AI factories and multi-megawatt deployments?
Yes. Modern liquid cooling systems—including scalable CDUs, end‑to‑end liquid loops, immersion‑ready designs, and chip‑to‑rack architectures—can scale from individual racks to multi-megawatt AI factories. This makes liquid cooling foundational for next-generation AI, HPC, and hyperscale deployments.
Related insights:
Liquid cooling lessons from HPC to AI factories: How to enable next-gen AI data centres