Microsoft plans to more than triple data center capacity by 2032, exceeding New York's power use
Editorial responsibility
- No named human review is recorded for this page.
- Source reporting is collected, normalized, translated or condensed automatically when needed.
- Automatically published source-backed update
Microsoft plans to more than triple the size of its global data centers by 2032, at which point their electricity consumption will exceed that of the entire state of New York, according to people familiar with the matter. The company's data center capacity will surpass 38 gigawatts in 2032, compared to New York State's projected summer peak demand of 33 gigawatts, representing growth from approximately 12 gigawatts currently. While market attention focuses on Nvidia chips for AI workloads, only about 2 gigawatts are currently allocated to AI-specific chips, expected to grow to about one-third of total capacity by 2032. The sources cautioned that plans may change due to shifts in customer preferences and new technologies. Microsoft initially declined to comment, then a spokesperson said the figures were inaccurate without providing alternatives, referring to the company's latest quarterly earnings report. Microsoft reported capital expenditures of $145 billion for its latest fiscal year, with analysts expecting continued growth. Investors reacted mutedly, with shares falling less than 0.5% in after-hours trading.
Source report
Source: Bloomberg
Microsoft plans to more than triple the size of its global data centers by 2032, at which point their electricity consumption will exceed that of the entire state of New York.
According to people familiar with the matter, Microsoft's data center capacity will surpass 38 gigawatts in 2032, compared to New York State's projected summer peak demand of 33 gigawatts for that year. This represents a growth of more than double from the company's current scale of approximately 12 gigawatts. One gigawatt is roughly equivalent to the power output of a nuclear reactor.
Although market attention focuses on Nvidia chips for artificial intelligence workloads, Microsoft's data centers remain primarily dedicated to general computing, with only about 2 gigawatts currently allocated to AI-specific chips. This figure is expected to grow to approximately one-third of the total by 2032.
The sources cautioned that these plans may change, as server farms require years to develop, and shifts in customer preferences and new technologies could disrupt underlying assumptions.
Company Response
Microsoft initially declined to comment on specific data center plans. After the report was published, a spokesperson stated that the figures were inaccurate but did not provide alternative numbers, referring instead to the company's most recent quarterly earnings report.
Microsoft reported capital expenditures (including finance leases) of $145 billion for its latest fiscal year, and analysts expect this spending to continue growing over the next few years. Most analysts have modeled forecasts only through 2029.
Market Reaction
Investors appeared to view the report as a signal that large-scale spending would persist, but reacted mutedly—Microsoft shares fell less than 0.5% in after-hours trading.
Related Quote
Nvidia CEO Jensen Huang said: "What better way to create demand than to manufacture problems? Who wouldn't want their market to be crazy about their products and lining up to buy them?" He made these remarks while promoting cybersecurity as the next major market for AI, following criticism that the technology has contributed to an increase in hacking incidents.
Other News
- Saudi Arabia's Public Investment Fund is considering merging Electronic Arts with Savvy Games Group
- Snapchat launches event planning features
- Oracle Cloud Infrastructure doubles quarterly revenue
- Elon Musk's Boring Company raises $3 billion at a $23 billion valuation, backed by investors including Abu Dhabi-based entities and Sequoia Capital
Original article: Bloomberg
Source
bloombergWestern