Microsoft’s Ambitious 38-Gigawatt Data Center Expansion by 2032

Microsoft is on a mission to triple its global data center capacity in what is being hailed as one of the most ambitious infrastructure buildouts in corporate history. A Bloomberg report reveals that the tech giant plans to expand its worldwide data center footprint from approximately 12 gigawatts today to a staggering 38 gigawatts by 2032. This more than threefold increase is driven almost entirely by a surging demand for AI computing power.

The decision to expand comes in the wake of severe hardware constraints that forced Microsoft to turn away paying cloud and AI customers. Notably, server shortages led e-commerce giant Temu to shift a major cloud deal to Oracle. In addition, GitHub experienced an eight-hour outage in August due to capacity constraints.

At present, Microsoft dedicates just 2 gigawatts to AI workloads. However, this share is projected to grow to roughly one-third of the 38-gigawatt total by 2032. This suggests a near six-fold increase in dedicated AI compute capacity.

In its latest fiscal year, Microsoft allocated $145 billion in capital expenditures, with $50 billion projected for Q1 of fiscal 2027 alone. Following the Bloomberg report, Nvidia shares climbed 0.5% in after-hours trading. Investors interpret Microsoft’s expansion as a sustained demand signal for AI chips.

However, analysts caution that grid connection queues lasting four to seven years and transformer manufacturing bottlenecks could complicate the execution of the plan.

Source: Yahoo Finance / GuruFocus — Microsoft Targets 38GW Data Center Buildout as AI Capacity Crunch Bites