Microsoft to triple data-center capacity to 38GW by 2032

Microsoft announced an ambitious data-center expansion roadmap that reflects the industry's intense infrastructure race to meet surging AI demand. According to Bloomberg reporting on September 10, 2026, the company plans to grow its owned-and-leased data-center footprint from approximately 12 gigawatts currently to more than 38 gigawatts by 2032—more than tripling its computing capacity. About one-third of the 26-gigawatt increase will be AI-specific hardware, growing from the current 2 gigawatts of AI-dedicated capacity. Microsoft has faced acute infrastructure constraints over the past year, forcing difficult choices around capacity allocation. Hardware shortages compelled the company to limit certain cloud subscriptions, turn away high-profile clients seeking AI services, and led to service disruptions affecting major platforms including GitHub and Xbox. The company's capital expenditure reached $145 billion in its most recent fiscal year, and it faces $329 billion in future data-center lease obligations. The scale of Microsoft's planned buildout is staggering: the full 38-gigawatt footprint would consume as much electricity as New York State during peak demand periods. New facilities under construction include an Atlanta-area cluster called East US 3, which is expected to bring 300 megawatts online this year and expand to over 1 gigawatt in coming years to relieve bottlenecks in Microsoft's largest data-center hub in Virginia. Sources cited by Bloomberg cautioned that the roadmap could shift as major data-center projects typically take years to develop and face growing regulatory and environmental scrutiny. The announcement underscores a broader pattern: every major cloud and AI company—including Alphabet, Amazon, Meta, and Oracle—is now in a sustained infrastructure expansion race, collectively earmarking trillions of dollars toward computing capacity to support next-generation AI systems.