A new analysis from Synergy Research Group reveals that just 20 state or metro markets worldwide now host 60 percent of all hyperscale data center capacity, underscoring the intense geographic concentration of cloud and AI infrastructure buildout.
Northern Virginia in the United States leads the pack, alone accounting for 11 percent of global hyperscale capacity. Beijing, China, follows as the second-largest market with a 6 percent share. The findings highlight how a handful of regions continue to dominate the location decisions of the world’s largest cloud and internet companies.
Among the top 20 markets, 15 are located in the United States. Four are in the Asia-Pacific region, including three in China—Beijing, Guangdong, and another unspecified market—with Singapore as the regional outlier. Europe is represented by just one entry: Dublin, Ireland.
The latest ranking reflects notable shifts. Tokyo, Japan; Sydney, Australia; and South Carolina have all dropped out of the top 20, replaced by emerging hubs such as Guangdong in China and the US states of Indiana and Tennessee. Beyond the top 20, the next 20 largest markets contribute an additional 19 percent of global hyperscale capacity, with several of those located in the Asia-Pacific region.
Synergy attributes the heavy US presence in the top 20 to the fact that 62 percent of hyperscale operators are headquartered in the country, and that the US generates nearly half of all cloud market revenue across key segments.
“A range of factors influence the choice of location for hyperscale infrastructure, including proximity to customers, availability and cost of real estate, availability and cost of power, networking infrastructure, ease of doing business, local financial incentives, political stability, and minimizing the impact of natural hazards,” said John Dinsdale, chief analyst at Synergy Research Group. “With the extremely rapid growth in demand for AI technology and infrastructure, availability of power has become an ever more critical criterion, as has the ability to overcome or work around local community objections to building large data centers. These factors are heavily influencing the geographic distribution of future infrastructure developments.”
Industry observers note that the findings carry significant implications for the data center sector. As AI workloads drive unprecedented demand for compute capacity, power availability and community acceptance are increasingly becoming decisive factors in site selection. The concentration of capacity in a small number of markets also raises questions about resilience, regulatory risk, and the long-term viability of secondary markets as potential growth areas.