Why Liquid and Hybrid Cooling are Powering the AI Boom

Artificial Intelligence (AI) is driving transformative change across both technology and the wider world. According to Gartner, worldwide spending on AI is forecast to reach $2.52 trillion in 2026, a 44% increase year-on-year.

However, as AI adoption expands, so does the thermal output of the hardware it relies on. This poses a growing challenge for data centre operators, who must ensure efficient thermal management to protect infrastructure and maintain performance, particularly as high-performance computing components, such GPUs generate significantly more heat than conventional IT equipment.  

Cooling costs present a major challenge in their own right. The Global Cooling Pledge, introduced at COP28, includes a commitment to reduce cooling‑related emissions by 68% by 2050. This has major implications for data centres, where an estimated 40% of total energy consumption is dedicated to cooling systems.

Traditional air‑based Cooling is Reaching its Limits

While conventional air‑cooling systems were designed for far lower rack densities, today’s compute requirements are pushing well beyond what air alone can effectively manage. AI workloads are already driving rack densities that exceed well beyond 100kW.

With overall data centre power demand expected to grow by 165% by 2030, this upward trend is set to continue. As a result, traditional air‑only cooling systems simply cannot keep pace.

How Liquid Cooling Drives Efficient, Sustainable Data Centres

Liquid cooling has emerged as an effective solution to tackle this challenge. Unlike traditional air-cooling systems, liquid cooling systems use fluids such as water or specialised coolants to absorb and transfer heat away from critical components. With far greater thermal capacity and conductivity than air, liquids can export heat at higher temperatures, enabling heat reuse and reduce the load on cooling systems. 

There are several types of liquid cooling technologies in use today. Some systems circulate coolant directly through racks to extract heat from servers, while others use cold plates attached to heat-generating components. Some go further still by fully submerging entire servers in thermally conductive, non-electrically conductive liquid, enabling comprehensive heat management across all components.

By targeting heat at its source, liquid cooling allows data centres to support higher compute densities and operate more efficiently, a critical capability as AI-driven demand for capacity continues to surge. 

Striking the Right Cooling Balance

Despite its advantages, liquid cooling is not a complete replacement for air cooling. Even with liquid systems in place, residual heat still escapes into the surrounding environment. As a result, air cooling is still required to maintain optimal conditions within the data hall.

For many data centres, a hybrid cooling strategy which combines liquid and air systems delivers the best results. The ratio of liquid-to-air will differ from customer-to-customer, but this approach ensures high thermal performance while also optimising Power Usage Effectiveness (PUE) and reducing energy consumption across the facility.  

Preparing for the Next Phase

As AI adoption accelerates, the pressure on data‑centre infrastructure will continue to intensify. Rising rack densities, expanding compute requirements, and increasing sustainability commitments mean that traditional air‑cooling systems can no longer support the demands of modern workloads. 

Liquid cooling offers the performance, efficiency, and environmental advantages needed to keep pace with this shift. But the future is not purely liquid. For most facilities, the most practical and scalable approach will be hybrid cooling architectures that combine the strengths of both technologies.

Ultimately, the organisations that embrace these next‑generation cooling strategies will be best positioned to unlock the full potential of AI. As compute demands rise and sustainability pressures mount, adopting liquid and hybrid cooling isn’t just a technical upgrade, it’s a foundational step in enabling the next era of high‑performance, sustainable data‑centre design.