Friday, June 26, 2026
EN·DarkSubscribe
AI Infrastructure · News & Analysis
HomeHeadlinesReport
Headlines · Report

AWS and Microsoft, the two dominant cloud providers, announce moderation in new data center deployment pace.

Slowdown from mega-scale players is a critical demand signal; indicates near-term capacity sufficiency concerns or ROI reassessment.
Trade pressSlicast · April 22, 2025 · Global · Source: pymnts.com
importance 85

AWS and Microsoft are pausing or slowing down the development of data centers, according to a Wells Fargo research note shared with PYMNTS. Microsoft said it remains on track to spend its 2025 budget of $80 billion or more for data centers but admitted to "slowing or pausing" early-stage projects based on "demand signals." AWS characterized changes to its data center plans as part of "routine capacity management."

Several industry sources told Wells Fargo's analysts that AWS has "paused a portion of its leading discussions" on colocated data centers, particularly those abroad. Noelle Walsh, president of Microsoft Cloud Operations and Innovation, disclosed that the software giant is "slowing or pausing some early-stage projects" as it refines cloud capacity buildout relative to customer demand. Microsoft has doubled its data center capacity in the last three years and now operates more than 350 data centers in at least 60 regions globally. Kevin Miller, vice president of global data centers at AWS, explained that the change represents "routine capacity management, and there haven't been any recent fundamental changes in our expansion plans," while noting AWS continues to see "strong demand for both generative AI and foundational workloads."

Wells Fargo analyst Eric Luebchow noted that "it's not clear yet" whether AWS' slowing of lease signing "is an area of concern or just the natural ebbs and flows of hyperscale activity." AWS has 114 availability zones and plans for 12 more in 36 global regions that serve 245 countries and territories. Meanwhile, other major data center owners — Meta, Google and Oracle — remain "active" while Nvidia is showing "elevated activity." The analysis excludes the $500 billion Stargate project to build artificial intelligence data centers involving OpenAI, SoftBank, Oracle and MGX, as well as Apple's data center plans, which are ramping up over the next four years.

Global data center capacity is expected to grow at 15% per year until 2027, which is forecast to be insufficient to meet growing demand, according to JLL. However, China's experience offers a cautionary tale. Following ChatGPT's launch in November 2022, more than 500 new data center projects were announced in China in 2023 and 2024, with at least 150 operational by the end of last year. Yet the rise of DeepSeek, which offered a more economical way to train and use AI models, combined with faltering economics around AI, has constrained the industry. Many data centers were constructed hastily and did not meet industry standards, leaving most companies running these facilities "struggling to stay afloat."

Read the original
AWS and Microsoft, the two dominant cloud… · Slicast