Data center operators face a critical challenge: expanding capacity to meet explosive AI demand while simultaneously reducing energy consumption and environmental impact. Rather than choosing between growth and sustainability, leading infrastructure providers are investing in targeted efficiency upgrades that lower electricity and water use across their facilities. These efforts signal a shift toward building AI-ready infrastructure that doesn't require proportional increases in power consumption.

What Are Data Centers Doing to Cut Energy Use?

Companies operating data centers are implementing a range of technical improvements designed to reduce energy consumption without sacrificing computing capacity. Flexential, a major data center operator, recently outlined its sustainability strategy in its 2025 ESG Report, detailing specific projects that demonstrate how efficiency gains are being achieved across the industry.

One of Flexential's largest initiatives involves upgrading air handling units across 14 data centers. The company estimates this program will reduce annual electricity consumption by 7.1 gigawatt-hours (GWh), a significant reduction that helps lower both operating costs and carbon emissions. In Las Vegas, a side-stream filtration system is expected to save 1.6 GWh of electricity annually while reducing water consumption by 2.6 million gallons per year. At the Salt Lake City facility, adiabatic precooling upgrades are projected to lower peak power usage effectiveness (PUE), a key efficiency metric, by 10.4% and reduce annual energy use by approximately 250 megawatt-hours.

How Are Cooling Systems Becoming More Water-Efficient?

Water management has emerged as a critical concern for data center operators, particularly as AI workloads intensify. Flexential reports that around 84% of its data center capacity now relies on closed-loop cooling technology, which continuously recirculates cooling water instead of drawing fresh water from external sources. This approach maintains reliable operations while dramatically reducing water consumption.

New facilities are being designed with even more ambitious targets. Flexential's newly developed sites are being built to achieve a Power Usage Effectiveness of 1.4, a metric that measures how much total facility power is needed to support actual computing equipment. The company is also targeting zero Water Usage Effectiveness through efficient cooling systems, meaning new facilities will recycle virtually all water used in the cooling process rather than consuming fresh supplies.

Steps to Evaluate Data Center Sustainability Performance

• Power Usage Effectiveness (PUE) Rating: Look for facilities with PUE scores of 1.4 or lower, which indicates that for every unit of power delivered to computing equipment, the facility uses minimal additional energy for cooling, power distribution, and other overhead systems.
• Renewable Energy Percentage: Check what portion of a data center's electricity comes from renewable sources. Flexential currently uses 402 gigawatt-hours of renewable energy across its operations, demonstrating the scale of clean energy integration possible in large facilities.
• Water Recycling Technology: Evaluate whether facilities use closed-loop cooling systems that recirculate water rather than consuming fresh supplies, and whether they employ additional conservation measures like side-stream filtration or rainwater harvesting.
• Facility Expansion Plans: Assess whether new capacity additions are being built to the same efficiency standards as existing facilities, ensuring that growth doesn't undermine sustainability progress.

Beyond data center infrastructure, technology companies are also addressing energy efficiency in their AI operations. Tesla's 2025 Impact Report highlights how waste heat recovery systems allow thermal energy generated by computing infrastructure to be reused within manufacturing operations, reducing overall energy demand. This approach demonstrates that efficiency gains can extend beyond cooling systems to capture and repurpose energy that would otherwise be wasted.

Why Are Companies Prioritizing Sustainability Alongside Expansion?

The shift toward efficiency-focused growth reflects changing customer expectations and business realities.

"Customers are increasingly evaluating infrastructure providers based not only on available capacity but also on sustainability and operational practices," said Ryan Mallory, CEO of Flexential.

Ryan Mallory, CEO of Flexential

This customer demand is driving infrastructure providers to embed sustainability into their expansion strategies rather than treating it as an afterthought. Flexential's expansion projects, including Portland-Hillsboro 4, which has entered service, and Portland-Hillsboro 5 and Atlanta-Douglasville 2, which are under development, are all being built to the same sustainability-focused design standards as the company's efficiency upgrade programs.

Governance and workforce development are also becoming central to sustainable expansion.

"We are enhancing the processes needed to manage growth, address emerging risks, and maintain reliable service for customers," explained Sam Rudek, Chief Operating Officer of Flexential.

Sam Rudek, Chief Operating Officer of Flexential

This emphasis on internal systems and oversight reflects recognition that sustainable infrastructure requires not just technical improvements but also organizational capabilities to manage complexity and accountability.

The broader technology industry is embracing similar principles. Tesla's sustainability strategy integrates renewable generation, battery storage, and electrified transport to support the transition to a lower-carbon energy system. The company is also investing in circular manufacturing practices, including water recycling technology, rainwater harvesting, and recovery of valuable materials from production waste and end-of-life batteries.

As artificial intelligence continues to drive demand for computing power, the infrastructure industry's focus on efficiency demonstrates that expansion and sustainability need not be opposing forces. By investing in advanced cooling technologies, renewable energy integration, and water conservation systems, data center operators are building the foundation for AI growth that doesn't require proportional increases in environmental impact.