The rapid expansion of AI data centres worldwide is driving a sharp increase in water consumption, raising concerns about environmental impact and water availability. Reports indicate that medium and large data centres use millions to billions of gallons of water annually for cooling, intensifying pressure in regions already facing water scarcity.
Medium-sized data centres use approximately 110 million gallons of water annually for cooling. This volume is sufficient to meet the yearly water needs of around 1,000 households.
Large data centres consume significantly more. According to reports, these facilities can use about 5 million gallons of water per day, amounting to approximately 1.8 billion gallons annually. This level of consumption is comparable to that of a small city.
AI data centres operate thousands of servers and high-performance GPUs that generate substantial heat. Without timely cooling, the risk of system failure increases.
To manage this heat, data centres rely on evaporative cooling technology, which uses water to regulate server temperatures. This method enables systems to operate continuously in a stable and efficient manner.
The increasing use of AI is driving a corresponding rise in water demand, raising concerns about its impact on environmental balance. The issue is particularly acute in regions already experiencing water shortages, where the presence of data centres may further strain limited resources.
Experts state that without timely adoption of alternative cooling technologies and improved resource management, the growing water consumption of AI data centres could become a major environmental challenge.
