At Mangum Economics, we leverage the best sources of information about data centers to quantify their impacts across North America.
The map shows our calculation of the megawatts of hyperscale and colocation data centers combined, based on information from datacenterHawk from the end of 2023. The shading is directly proportional to the operational critical capacity at the time.
We separated Northern Virginia from the rest of the state. The region alone had as much critical capacity as all of Oregon, Texas, Arizona, Nebraska, and Pennsylvania combined. Northern Virginia is roughly balanced between hyperscale and colocation capacity.
Oregon’s second-place ranking (with 38 percent of Northern Virginia’s capacity) was largely due to hyperscale development around Boardman and Prineville.
Texas was in third place (with 33 percent of Northern Virginia’s capacity) spread across Dallas / Fort Worth, Houston, and Austin (with mostly or exclusively colocation development) and San Antonio (with mostly hyperscale capacity).
Behind those top three regions, the rest of the top ten regions (Arizona, California, Iowa, Ohio, Illinois, Georgia, and the rest of Virginia) were roughly the same size. The data center investment in Iowa and Ohio was almost entirely from hyperscale development. Hyperscale development makes Virginia a top-ten state even without Northern Virginia.
The rest of the shaded states had roughly the same amount of critical capacity, and in most of them, hyperscale investment accounted for almost all of the data center development.
The high rankings of Oregon, Iowa, Ohio, and the rest of Virginia become apparent only after accounting for hyperscale development. Because most of the AI investment is being done by hyperscalers, and hyperscale development attracts colocation development, we expect more intensive development in Oregon, Nebraska, Iowa, Ohio, the rest of Virginia, and the Carolinas.
ABOUT THE AUTHOR
David Zorn, Director of Technology & Special Projects, based in Northern Virginia. He previously spent over 20 years at the U.S. Food and Drug Administration, where he ultimately served as the Director of Social Sciences leading research on consumer responses to labeling, public health messaging, and the economic impacts of FDA policies. He has published papers on quantitative risk assessment and data quality. For Mangum Economics, David has been the primary author of scores of data center impact analyses across the United States and Mexico. His reports show how data centers benefit state and local economies. He has extensive knowledge of and contacts in the industry that are essential at keeping the data center analyses current and accurate.
