Meta’s founder, Mark Zuckerberg, announced that the social media giant would be spending hundreds of billions of dollars to build massive AI data centres in the US. He revealed that the first multi-gigawatt data centre, named Prometheus, was expected to come online in 2026.
According to him, one of the planned sites would cover an area almost as large as Manhattan, which spans 59.1 square kilometres (22.8 square miles).
Meta has been investing heavily in the development of what it refers to as “superintelligence” — technology that it believes could surpass the cognitive abilities of the smartest humans.
The company, which primarily generates its revenue from online advertising and earned over $160 billion in 2024 alone, is planning to build several multi-gigawatt clusters. Zuckerberg stated in a post on his social media platform, Threads, that one of these clusters, called Hyperion, could scale up to five gigawatts over the coming years.
“We’re building multiple more titan clusters as well. Just one of these covers a significant part of the footprint of Manhattan,” he added.
Prometheus will be built in New Albany, Ohio, while Hyperion will be built in Louisiana and is expected to be fully online by 2030, Zuckerberg said.
Read also: Meta Tried To Get OpenAI Staff With $100m Offers – Altman
Mark Zuckerberg said that Meta would be investing hundreds of billions of dollars to build superintelligence and that the new data centres had been given names that reflected their immense scale and impact.
Karl Freund, principal analyst at Cambrian AI Research, told the press that it was clear Zuckerberg intended to spend his way to the top of the AI industry. He added that the talent Meta was hiring would have access to some of the best AI hardware in the world.
According to Reuters, Meta’s shares were trading 1% higher following the announcement, and the stock had already risen by more than 20% so far this year.
Globally, there are at least 10,000 data centres hosting the cloud, with the majority located in the US, followed by the UK and Germany.
However, it was noted that AI-driven data centres consume enormous amounts of energy and water. One study estimated that such centres could use up to 1.7 trillion gallons of water globally by 2027. It was also reported that a single AI query — such as a request to ChatGPT — could consume about as much water as a small bottled drink from a local shop.