Building the infrastructure behind a more digital world

AI may be transforming the economy digitally, but its growth ultimately depends on physical infrastructure. Water infrastructure plays a critical role in enabling that growth by helping data centers manage the increasing demands of cooling, energy and computing capacity efficiently and reliably. As AI drives greater computing density and data center expansion, taking a smarter, system-level approach to water will be essential to supporting digital growth while managing resources responsibly. 

According to the International Energy Agency¹, global electricity consumption from data centers is projected to nearly double between 2025 and 2030, from about 485 to 950 terawatt-hours, driven in large part by the growth of AI.

That scale of growth brings tremendous opportunity. It also brings real challenges. More computing power generates more heat. More heat requires more cooling. And as data centers expand, communities are asking important questions about the energy and water required to support them.

The question isn’t whether digital infrastructure will continue to grow. It’s how we help that growth happen as efficiently, reliably and resiliently as possible, without putting unsustainable pressure on the water systems and other resources communities depend on.

For those of us in the water industry, that is a challenge we can help solve.

Growth and efficiency have to happen together

According to GWI and Xylem’s Watering the New Economy report², the new economy withdrew 23.7 trillion liters of water in 2025, an increase of 38% from 2020. As digital infrastructure continues to expand, total water demand is expected to increase as well.

At the same time, technology is becoming more efficient. The report projects that data centers will become 46% more water-efficient by 2050, saving an estimated 2.4 trillion liters of water annually.

Demand and efficiency will rise together. Even as facilities become significantly more water efficient, overall resource requirements are expected to increase as the industry expands. That makes efficiency an essential component.

Water and energy are part of the same system

It’s easy to think about a data center’s water footprint only in terms of the water used at the facility. The reality is more complex.

Advanced cooling technologies are changing the relationship between computing growth and water demand. Closed-loop cooling systems recirculate cooling fluids within sealed loops, reducing reliance on water-intensive cooling processes and lowering freshwater demand. At the same time, liquid cooling and other advanced thermal management technologies are enabling more efficient heat removal, helping data centers support growing computing loads with a smaller water footprint.

But onsite use is only part of the picture. Electricity generation also has a water footprint. Watering the New Economy projects that by 2030, power supply will account for 48% of total data center water consumption.

This connection between water and energy reinforces why cooling efficiency matters. Cooling systems require electricity to operate, and the energy used to power them can carry its own footprint. Improving cooling performance offers an opportunity to address both energy demand and the broader water impact of data center operations.

Think beyond the cooling system

Behind every data center is an interconnected infrastructure system working to manage heat, water and energy as computing loads change. Pumps, heat exchangers, tanks and controls all play a role, but optimizing individual components is only part of the challenge.

As computing density increases, operators and designers need to think about water and cooling infrastructure as a complete system. That means understanding where water will come from and how much will be needed, while considering how efficiently cooling liquids circulate, how systems respond to changing workloads and where water losses can be reduced. The goal is not simply to meet today's cooling requirements, but to create infrastructure that can balance efficiency, reliability and performance as demand grows.

At Xylem, we've spent decades helping customers move, manage and optimize water in some of the world's most demanding environments. That experience is increasingly relevant to data centers, where decisions about water sourcing, circulation, heat transfer, controls and reuse are becoming interconnected. These challenges are connected. Compact, high-performing equipment can give designers greater flexibility in how mechanical spaces are configured, while controls can help systems respond dynamically to changing conditions. Together, these capabilities can help operators maintain reliable cooling performance as computing loads increase and available space becomes more constrained.

As cooling requirements become more complex, customers need partners who understand not only individual technologies, but how the broader water and cooling infrastructure work together. Taking that system-level view can help data centers support higher computing densities while managing water and energy more efficiently and building greater resilience into operations.

Efficiency isn’t enough

Data center cooling also has to be reliable and resilient. Cooling is mission-critical infrastructure. Customers need systems that can perform consistently as loads change and facilities scale. And increasingly, they need to do it within a constrained physical footprint.

The technology will continue to change, too. AI is increasing computing density, while liquid cooling and other approaches are changing where and how heat is managed. But one fundamental need remains: heat has to be moved somewhere.

The architecture may change, but the need to circulate cooling fluids, transfer heat and manage resources efficiently and reliably isn’t going away.

Helping growth happen better

Water is only one part of the broader conversation communities are having about data center growth. Power availability, grid capacity, noise, land use, jobs and other local impacts matter, too.

Those concerns shouldn’t be minimized. But neither should the opportunity to help solve them.

Across Xylem, we’re seeing firsthand how quickly the data center market is growing and how important water and cooling infrastructure will be to what comes next. These challenges sit at the intersection of water, energy efficiency, system performance and resilience, areas where we’ve been helping customers solve complex challenges for decades.

For us, that means working alongside customers to design and optimize water infrastructure that is efficient, reliable and resilient. By bringing together expertise across the water cycle, from sourcing and treatment to cooling and reuse, we can help customers and communities meet growing digital demand while managing water resources responsibly.

Frequently asked questions about water and data centers

Why is water infrastructure important to AI and data center growth? AI is increasing computing density and the amount of heat that data centers must manage. Water and cooling infrastructure help move that heat efficiently and reliably, making them an important part of the physical infrastructure behind digital growth. Taking a system-level approach can also help operators balance cooling performance with water use, energy demand and long-term resilience.

Why does AI growth increase water demand at data centers? Greater computing demand generates more heat, which increases cooling requirements. Data center water demand can also extend beyond water used directly at the facility because electricity generation can have its own water footprint. As AI-driven computing expands, considering both direct and indirect water use becomes increasingly important.

How can data centers improve water efficiency? Data centers can improve water efficiency by looking beyond individual pieces of equipment and considering the performance of the entire cooling system. Closed-loop cooling can reduce reliance on water-intensive processes, while advanced thermal management, controls and more efficient circulation can help facilities manage growing computing loads with a smaller water footprint.

How can data centers balance water efficiency with reliability? Efficiency and reliability need to be considered together because cooling is mission-critical infrastructure. A system-level approach can help operators understand how cooling liquids are circulated, how controls respond as workloads change, where losses can be reduced and how infrastructure can accommodate increasing computing density while maintaining consistent performance.

Digital infrastructure will continue to grow. The opportunity isn't simply to build more of it, but to build it smarter, with efficiency, resilience and responsible resource use designed in from the start.

^{1 International Energy Agency https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai
2 Watering the New Economy}