As the race for A.I. dominance continues, companies like Google (GOOGL), Meta (META), OpenAI, Microsoft (MSFT) and now Elon Musk’s xAI are investing more money, time and resources into securing and developing data centers across the U.S. and the globe. Just this week, xAI announced plans to build a data center in Memphis, Tenn. in a former manufacturing facility—a move that would bring significantly more computing capacity to train the company’s A.I. models that power its ChatGPT rival, Grok. The announcement came weeks after xAI closed a $6 billion venture funding round, which valued the just year-old startup at $24 billion.
Data centers are critical to the development and training of any A.I. language model, as its operation requires a large central source for electricity, storage and cloud computing. Data centers resemble “A.I. factories,” the Wall Street Journal opined, citing Nvidia CEO Jensen Huang, who has said data centers are an essential part of the “next Industrial Revolution.” Indeed, many data centers in the U.S. have had a significant impact on the economies of the cities they reside in—adding hundreds of jobs to the local economy. For example, Microsoft’s network of data centers in Iowa employ over 300 West De Moines residents, making it among the 10 largest employers in the city of over 70,000.
For now, xAI’s plans to develop a data center in Memphis are still pending approval by local and regional authorities—but the project seems to be moving quickly. Memphis Chamber of Commerce President Ted Townsend told the Wall Street Journal that an inquiry from the company came in March. If approved, the project could break ground this year.
It’s possible that Musk chose Memphis as the home for xAI’s new data center because of a potential for tax breaks, although none such incentives has been confirmed by the Memphis and Shelby County Economic Development Growth Engine (EDGE), the area’s regional department overseeing tax incentives for businesses.
Where are the other A.I. juggernauts keeping their data?
Google, Meta, Microsoft, Amazon (AMZN) and OpenAI likely operate under similar motivations, all having poured billions of dollars into data centers throughout the U.S. Criteria may also include land availability, skilled workforce population, energy resources, and other infrastructural needs.
Microsoft and OpenAI: OpenAI has not disclosed the exact location of its data centers, but has partnered with Microsoft in the past to distribute servers worldwide on Amazon’s Azure platform, including a collaborative custom computing system in Iowa used to train large-scale A.I. models. In addition, Microsoft has data centers in Virginia, Arizona, Texas, Washington D.C. and California.
Amazon Web Services (AWS) has data centers worldwide to power its cloud services and supercomputing infrastructure, with locations in Virginia, Seattle, California and Oregon, as well as internationally in Ireland, Luxembourg, Germany, China, Japan, Singapore, Australia and Brazil.
Google: Google has U.S. campuses in North and South Carolina, Ohio, Iowa, Oregon, Georgia, Texas, Nevada, Alabama, Oklahoma, Tennessee and Nebraska. The company also operates data centers internationally and has more under development.
Meta: Meta’s data centers spread across the country, with 20 locations currently, as well as a few international ones (and likely more to come, as mapped on the company’s website).