The Role of Coolant Distribution Units in High-Density Data Centres

Artificial intelligence (AI) continues to redefine the boundaries of computing performance. From platforms like ChatGPT to emerging AI-native applications, the demand for next-generation processing is accelerating faster than ever before. This surge is driven by the need for high-performance CPUs and GPUs, whose thermal design power (TDP) is reaching unprecedented levels that challenge traditional cooling systems.

That’s where liquid cooling comes into play. As rack densities rise with accelerated computing, direct liquid cooling (DLC)—enabled by CDUs—is projected to grow from $1.1B in 2024 to $5.8B by 2029, leading innovation in the data center cooling market. In the AI era, thermal management has evolved into a strategic enabler of high-performance computing. As processors continue to push the limits of power density, only advanced thermal solutions can ensure sustained performance, reliability, and efficiency required to support high-intensity AI workloads. 

Cooling Trends in AI Data Centers

Liquid cooling is rapidly transitioning from a ‘nice to have’ to a necessity. In 2023, only 10% of data centers deployed liquid cooling, but by 2030, that number is projected to reach 50%. This is a result of the escalating thermal demands of AI and high-performance computing (HPC), as well as the growing pressure to meet sustainability goals. Traditional air cooling systems can consume up to 40% of a data center’s energy, while advanced cooling solutions aim to reduce that figure to as little as 5% of the total IT load - significantly improving power usage effectiveness.

Water conservation is also becoming a critical focus. According to data from the Financial Times, data center water consumption has surged from 1.13 billion to 1.85 billion liters between 2019 and 2023—an increase of nearly 64%. In response, technologies that reduce or eliminate evaporative cooling are gaining traction. Closed-loop systems and eco-friendly coolants are helping operators reduce their environmental footprint while maintaining performance.

Understanding Coolant Distribution Units in Data Center Cooling Systems

A Coolant Distribution Unit (CDU) is a centralized control system in liquid cooling architecture, responsible for managing the flow of coolant between IT equipment and the facility’s water system (FWS). It regulates temperature, flow rate, and pressure, while also monitoring system health, detecting leaks, and enabling remote diagnostics. This makes CDUs essential for maintaining stable thermal conditions in high-density, scalable computing environments.

According to the ASHRAE TC 9.9 Technical Alert, effective liquid cooling benefits from thermal isolation between the Technology Cooling System (TCS) and the facility water system. CDUs are central to this architecture, acting as the interface that separates these two loops and manages the TCS loop. This isolation improves control, can prevent condensation, and allows for precise management of coolant temperature, flow, and quality. It also protects sensitive IT hardware from exposure to untreated facility water, which may contain minerals, particulates, and other impurities that could compromise system integrity.

In operation, CDUs circulate coolant through cold plates or other heat exchangers mounted on processors or GPUs. After absorbing heat, the warmed coolant returns to the CDU, where it either transfers that heat to the facility water loop or dissipates it via an air-cooled radiator. This closed-loop process supports higher compute densities while reducing reliance on traditional air cooling.

Beyond performance, CDUs play a critical role in enhancing cooling system resiliency. According to a recent Uptime Institute study, over 70% of unplanned downtime in data centers is linked to power and cooling system failures. By enabling precise thermal control and proactive fault detection, CDUs help operators reduce downtime risk, maintain service continuity, and ensure reliable operations—even under the intense demands of AI and HPC workloads.

Choosing the Right CDU for Your Data Center Cooling Strategy

Selecting the appropriate CDU is a critical step in designing an effective liquid cooling strategy. The choice typically comes down to two main types: Liquid-to-Liquid and Liquid-to-Air CDUs, each suited to different infrastructure environments and performance goals.

Liquid-to-Liquid (L2L) CDUs are designed for facilities with access to a dedicated facility water system. These systems offer superior cooling capacity and efficiency by leveraging water’s high thermal conductivity. Heat is transferred from the IT coolant loop to the facility water loop via an integrated heat exchanger, making this approach ideal for high-density deployments with significant thermal loads. 

In contrast, Liquid-to-Air (L2A) CDUs use air-cooled radiators and fan systems to dissipate heat into the surrounding environment. This makes them well-suited for locations without facility water access or for pilot deployments exploring liquid cooling. While they offer greater flexibility in terms of installation, they typically deliver lower cooling efficiency and ultimately still require heat loads to be removed from the air at the facility level.

Though both CDU types share core components—such as pumps, filtration systems, and temperature regulation—their heat rejection methods and operational considerations differ significantly. The decision ultimately hinges on infrastructure availability, workload intensity, and long-term scalability goals. For environments prioritizing high-efficiency cooling and growth, L2L CDUs are the preferred solution. For more flexible or transitional deployments, L2A CDUs present a practical alternative, offering a low-friction pilot path to liquid cooling.

Future-Proofing AI and HPC Environments with CDUs

As AI and HPC workloads continue to grow in complexity and power density, cooling infrastructure must keep pace. Single-phase direct-to-chip liquid cooling, delivered through CDUs, offers a practical path forward, helping stabilize operating temperatures while reducing energy use, supporting heat reuse, and minimizing water consumption.

These systems are helping to define what efficient, reliable, and climate-conscious data center operations will look like in the years ahead.