Petroc Taylor
Research expert covering global developments in the use of data
Data centers - statistics & facts
Choose a region:
Data centers lie at the heart of the AI revolution. Housing IT equipment organized into racks, data centers are powering AI, the cloud, and other data-intensive technologies by efficiently processing workloads and storing data securely. Once used to describe small, on-site server rooms for businesses, the term 'data center' now refers to vast, power-intensive facilities operated by global tech giants. Colocation operators such as Equinix and Digital Realty offer data center capacity for rent at shared facilities across the globe, while big tech firms such as Meta and Google operate their own hyperscale facilities to power cloud and AI offerings. With data center demand surging across the globe, governments and regulators are grappling with how to nurture this essential digital infrastructure while addressing sustainability concerns.
Key markets feel the strain
Major data center operators have typically looked to cluster facilities in established hubs with favorable local infrastructure and proximity to major population centers. The Northern Virginia region is undoubtedly the world’s largest single data center hub, offering the most inventory of any region, while the FLAPD (Frankfurt, London, Amsterdam, Paris, and Dublin) markets serve as key European hubs. However, with global data center supply projected to grow by more than 20 percent in 2025 alone, the pace and scale of expansion has prompted both operators and local officials in key regions to reassess. A major issue is sourcing power, with the industry placing a massive strain on local power grids. In Ireland, for example, data centers consume over 20 percent of the country's electricity, while in Virginia, the figure exceeds 25 percent.Water cooling compounds environmental concerns
Energy consumption—and sustainability more broadly—will be a defining challenge as demand continues to surge. Data center operators were previously able to offset rising demand through relatively low-cost efficiency measures, such as basic cooling upgrades. However, further gains have become increasingly costly and complex to achieve. At the same time, the adoption of modern GPUs built to support intensive AI workloads has introduced new challenges. These AI-specific GPUs are designed to operate at higher temperatures, necessitating the use of intricate water-cooling systems. As a result, water consumption has risen sharply in modern facilities, raising tensions in regions negotiating water scarcity.The EU seeks to increase transparency
While awareness is growing around the substantial environmental cost of the data center industry, a lack of transparency makes it difficult for policymakers and the wider public to grasp the true scale of the challenge. The European Union has sought to tackle this with new legislation requiring large operators to report key metrics. Regardless, government leaders are aware that expanding digital infrastructure is necessary to unleash the economic potential of the AI revolution, making the path forward a balancing act between regulation and growth.Petroc Taylor
,
Apr 24, 2025
Petroc Taylor
Research expert covering global developments in the use of data
