Why data centers are moving to liquid cooling

Modern data centers are experiencing a surge in computing demands, largely driven by advancements in artificial intelligence (AI) and high-performance computing (HPC). Serious “gamers” have been installing liquid cooled processor cooling in gaming computers for years. In fact, liquid chip cooling in data centers is not new, it has been applied to high demand systems for more than a decade. 

The demands of AI applications necessitate more powerful processors that generate significant heat, exceeding the capabilities of traditional air-cooling methods. Direct-to-chip liquid cooling is becoming an increasingly important technology, offering significant advantages in efficiency, performance and sustainability.

Key benefits of direct-to-chip liquid cooling

 Higher heat transfer where it matters most

Aqueous glycols, as we know from other cooling applications, offer significantly higher thermal conductivity and heat capacity compared to air. This enables much more efficient heat removal from CPUs, GPUs and other critical server components, helping maintain stable operating temperatures even under demanding AI and HPC workloads.

 Reducing cooling energy demand

Liquid cooling can significantly reduce the energy required to remove heat compared with traditional air-cooled systems that rely on large numbers of fans and air-conditioning units. As data centers work towards Net Zero and other sustainability targets, improving cooling efficiency can have a meaningful impact on both energy consumption and operating costs. Some estimates suggest that liquid cooling can reduce cooling-related energy costs by more than 50%.

 Keeping CPUs and GPUs operating at peak performance

Liquid cooling helps prevent thermal throttling—a built-in safety mechanism that automatically reduces processor performance when temperatures become too high. To protect the hardware from overheating, CPUs and GPUs lower their clock speed—the rate at which they execute instructions. While this prevents damage, it also means workloads take longer to complete and overall computing performance is reduced.

By efficiently removing heat directly from the processor, liquid cooling helps maintain stable operating temperatures, allowing CPUs and GPUs to sustain higher clock speeds and peak performance, even during demanding AI and HPC workloads. This also contributes to improved hardware reliability and a longer service life, reducing the likelihood of component failures and replacements.

 Supporting higher rack densities

As computing demand continues to increase, operators are looking to deploy more processing power within existing data center footprints. Thanks to its superior heat transfer capabilities, direct-to-chip liquid cooling enables higher rack densities than conventional air cooling. Dell highlighted this trend, noting that liquid cooling is becoming an important enabler for high-density AI infrastructure where physical space is limited.

Why coolant selection matters in liquid cooling

The benefits of direct-to-chip liquid cooling depend not only on the cooling technology itself but also on the coolant circulating through the system. Thermal performance, corrosion protection, material compatibility and long-term fluid stability all play a role in maintaining reliable and efficient operation throughout the lifetime of the installation.

ZITREC^® EC for direct-to-chip liquid cooling

ZITREC^® EC liquid cooling addresses the needs of AI processing in datacenter cooling, offering advantages over air-cooling. By enhancing cooling efficiency, reducing energy consumption and costs, improving performance and reliability, and optimizing space utilization, ZITREC^® EC cooling is an essential component for modern data centers aiming to meet the escalating demands.

To support customers' sustainability ambitions, ZITREC^® EC is also available as an ECO variant with a significantly reduced Product Carbon Footprint (PCF) compared with its virgin fossil-based equivalent, while delivering the same quality and performance. This is achieved by using base fluids derived from bio-based or recycled raw materials, allocated through a certified mass balance approach.