A recent Washington Post article highlighted a growing and urgent issue in the tech industry: the staggering water consumption of data centers, particularly those deploying artificial intelligence (AI) workloads. As the article explains, the unprecedented computational demands of AI are dramatically increasing both electricity and water usage, straining resources and raising environmental concerns. However, there is a solution to the water crisis in data centers—liquid cooling.

As the article rightfully noted, much of the water used in data centers is for cooling, particularly through evaporative cooling systems. In these systems, heat generated by computer chips is removed via air cooling, which blows hot air over chillers. The chillers then use evaporating water to release heat into the atmosphere, consuming billions of gallons annually.

With the rapid rise of AI and other high-powered compute applications, the heat emitted by AI chips has increased drastically, forcing data centers to use even more water to maintain temperatures. The article emphasizes how this trend is unsustainable, especially as AI continues to expand its footprint across industries.

Fortunately, there is a way to dramatically reduce the massive water consumption in data centers: liquid cooling. Unlike traditional air-based systems, liquid cooling transfers heat directly from the hardware using liquids, which are much more efficient at capturing and dissipating heat. Direct heat removal with liquid decreases or even eliminates the need for water evaporation, offering a water-saving alternative that is perfectly suited to handle the heat generated by AI workloads.

In fact, liquid cooling can go beyond just cooling—it can repurpose the heat for secondary uses, such as heating nearby buildings, supporting agriculture, or even aquaculture projects. Instead of simply wasting heat and water, liquid cooling creates a closed-loop system that reduces overall resource consumption and environmental impact.

AI is driving the adoption of liquid cooling

As the Washington Post article states, AI is accelerating the need for change. While liquid cooling has been a long-standing option, it has often been sidelined in favor of conventional air cooling. However, with AI pushing the boundaries of what data centers can handle, liquid cooling is no longer a secondary option—it’s becoming a necessity.

Nvidia’s recent announcement that its new Blackwell architecture, designed specifically for AI, will be entirely liquid-cooled is a clear signal that the industry is beginning to understand this shift. With the rising heat intensity of AI chips, air cooling is hitting its limits, and the water consumption tied to traditional evaporative cooling has skyrocketed. The future is clear: AI will drive the widespread adoption of liquid cooling.

In addition to reducing water usage, liquid cooling offers a major boost to energy efficiency. As the article notes, AI’s rising energy demands are a concern. Data centers traditionally dedicate up to 40 percent of their energy consumption to running fans and chillers in air-cooling systems. Liquid cooling slashes this energy use by providing a more direct, efficient cooling mechanism, reducing both operational costs and the carbon footprint of data centers.

As AI continues to push the limits of compute power, the need for sustainable, resource-efficient cooling solutions has never been greater. The Washington Post article highlights the growing energy and water demands, but one of the solutions is already here—liquid cooling. Fortunately, some innovative new data center developers like Wyoming Hyperscale, Ada Infrastructure, and Start Campus have embraced the benefits of liquid cooling by designing their data centers to operate using liquid cooling from the start.

Many existing data centers are also working on liquid cooling retrofits for at least parts of their data centers. The challenge now is scaling this technology as rapidly as possible for all data centers interested in deploying AI and high-powered compute applications.

As AI continues its relentless growth, we need to ensure that the data centers powering this revolution evolve as well. Liquid cooling is a key to solving both the water and energy challenges posed by AI. The time to act is now.