Meta is pouring hundreds of billions into massive AI data centres in the US.
The first mega centre, dubbed Prometheus, will hit New Albany, Ohio by 2026. It’s a “multi-gigawatt” beast—one site alone will span nearly the size of Manhattan (22.8 square miles).
Mark Zuckerberg also announced Hyperion, another multi-gigawatt cluster in Louisiana, set to go fully live by 2030. This cluster could scale up to five gigawatts over several years. 
Zuckerberg said Meta will build multiple “titan clusters” with names to match their mammoth scale and impact.
The move backs Meta’s push toward “superintelligence,” AI tech designed to outthink top human minds. Despite mostly making money from online ads, Meta raked in over $160 billion in revenue in 2024.
Karl Freund, principal analyst at Cambrian AI Research, told the BBC Zuckerberg is ready to “spend his way to the top of the AI heap.”
“Clearly, Zuckerberg intends to spend his way to the top of the AI heap.”
“The talent he is hiring will have access to some of the best AI Hardware in the world.”
Karl Freund, Cambrian AI Research
Meta’s stock was up 1% right after the announcement. Shares have climbed over 20% this year.
There are 10,000+ data centres worldwide hosting cloud services, mostly in the US, UK, and Germany.
AI data centres guzzle energy and water. One study warns global consumption could reach 1.7 trillion gallons of water by 2027. A single AI query, like asking ChatGPT, uses about as much water as a small bottled drink.
Meta is betting big on AI hardware and infrastructure in the US, aiming for dominance in next-gen AI technology.
