The United States is in the middle of the largest infrastructure buildout since the interstate highway system. 6 hyperscalers — Amazon, Google, Meta, Microsoft, Oracle, and the Stargate consortium — have committed over $690 billion in capital expenditures to build AI-ready data centers across the country. 74 new facilities broke ground in 2026 alone, spanning 28 states. This tracker monitors every major project, from land acquisitions and power contracts to construction milestones and GPU deployments, giving investors and policymakers a real-time view of where the AI economy is being physically built.
$690 billion in cumulative CapEx has been committed by the six largest hyperscalers — Amazon, Google, Meta, Microsoft, Oracle, and the Stargate consortium. Microsoft alone has pledged $80B for fiscal 2025, while Amazon Web Services is investing $100B across new US regions through 2028.
4,000+ data centers currently operate across the US, with 74 new AI-focused facilities breaking ground in 2026 alone. Virginia’s Data Center Alley in Northern Virginia remains the densest cluster, hosting roughly 35% of total US capacity by power draw.
4.7 million construction and operations jobs are projected through 2030 from AI infrastructure projects. Each large-scale data center campus creates 3,000-7,000 construction jobs over a 2-3 year build cycle, plus 200-500 permanent operations roles.
$500 billion has been earmarked for Project Stargate, a joint venture between OpenAI, SoftBank, Oracle, and MGX to build the largest AI data center complex in history. The first campus in Abilene, Texas broke ground in early 2025 with a planned 1 GW initial capacity.
37 gigawatts of total US data center power demand is projected by 2030, up from roughly 17 GW in 2023. AI workloads are the primary driver, with a single large GPU cluster consuming 50-100 MW — equivalent to powering 40,000 homes.