EXECUTIVE SUMMARY:
- Cooling has gone from technical consideration to strategic enabler, with the rate of data centre expansion dictated by how effectively they manage heat
- Liquid cooling meets demands for edge computing as well as increasing sustainability challenges with technical efficiency, discreet operation, and lower energy consumption
- Operators should treat liquid cooling as a foundational infrastructure and platform-based solution to ensure consistency and resilience as AI evolves
AI is rewriting the rules of digital infrastructure, driving an unprecedented surge in energy demand as each query, output, and generative task triggers billions of calculations. The International Energy Agency projects that data centre consumption could reach 1000TWh by 2030, more than double today’s levels while hyperscale facilities now draw hundreds of megawatts. NVIDIA forecasts racks could hit 1MW apiece by 2027. With every watt consumed by processors turning into heat, cooling already accounts for almost 40% of total data centre energy use. The trajectory shows that AI is not only accelerating digital innovation but reshaping the physical and environmental boundaries of compute itself.
For over three decades, air cooling has moved from niche experiment to proven necessity. The Uptime Institute reports that nearly a quarter of operators have already adopted direct liquid cooling, with uptake accelerating as AI workloads scale to industrial levels.
Cooling has evolved from a technical consideration to a strategic enabler. The rate at which data centres can expand is now dictated by how effectively they manage heat. Future-proofing infrastructure requires designs capable of supporting densities once deemed impossible, with cooling systems that are modular, globally deployable, and as serviceable as the compute they protect. In AI factories and edge installations alike, downtime is simply not an option.
The conversation extends beyond hyperscale environments, with edge computing poised for explosive growth as enterprises deploy AI inference closer to users where latency is critical. Smart cities, autonomous vehicles, and healthcare applications will increasingly rely on compute located in dispersed, smaller facilities – often in urban or resource-constrained settings. These locations introduce unique pressures: limited access to power, strict noise regulations, and growing community concerns about sustainability. Liquid cooling meets these demands by offering not only technical efficiency but also discreet operation, lower energy consumption, and reduced dependence on evaporative water systems. In edge deployments where space is tight and serviceability is paramount, modular liquid cooling platforms are essential.
Sustainability pressures, meanwhile, will only intensify. Cooling is no longer regarded merely as an operational challenge: it is a matter of climate responsibility. Policymakers in Europe, Asia, and North America are sharpening their focus on emissions, water usage, and energy efficiency in digital infrastructure with communities increasingly resisting new builds that lack credible sustainability strategies, forcing operators to demonstrate not just capacity but accountability. Liquid cooling provides a direct response by reducing overall power usage, limiting water draw, and enabling heat reuse initiatives that benefit local communities.
Viewing liquid cooling as a platform, rather than a retrofit utility, completes the picture. Treating it as foundational infrastructure allows operators to scale without added complexity. Platform-based solutions ensure consistency across new builds and retrofits, support sustainability objectives, and bolster resilience amid regulatory and community pressures. This matters because AI’s future will be defined as much by environmental responsibility as by technical possibility.
The coming decade will bring heightened challenges. Thermal loads reaching extremes, governments demanding stricter safeguards, and communities expecting greater accountability will ultimately mean data centres that fail to adapt will be constrained by both physics and politics. Those that embrace liquid cooling will have the means to scale responsibly, balancing ambition with sustainability. More than just an efficiency measure it becomes a licence to operate in a world where digital growth must coexist with environmental limits.
AI may be driving the next industrial revolution, but liquid cooling will ensure this revolution is sustainable. The question is no longer whether it is necessary – it’s how rapidly it can be deployed as the backbone of tomorrow’s digital economy.
About the author: Stuart Crump is Global Commercial Director at LiquidStack.
