Meta Bets Hundreds of Billions on "Manhattan-Sized" AI Supercomputers
CEO Mark Zuckerberg has unveiled plans for Prometheus and Hyperion, multi-gigawatt AI data centers designed to dominate the race for artificial general intelligence.
Olivia Sharp
•
•
1 min read
•
411 views
Free
Meta is investing "hundreds of billions" to build the world's most powerful AI supercomputers, including the 1-gigawatt Prometheus and the 5-gigawatt "Manhattan-sized" Hyperion, in a massive bid to win the race to AGI.
Meta's "Manhattan Project" for AGI
Meta is going all-in on the AI infrastructure arms race. On July 14–15, CEO Mark Zuckerberg announced a plan to invest "hundreds of billions of dollars" to build the world's most powerful AI supercomputing clusters—a move aimed squarely at achieving artificial general intelligence (AGI).
The Superclusters: Prometheus and Hyperion
Meta's plan involves facilities on an unprecedented scale:
- Prometheus: A 1-gigawatt data center coming online in Ohio in 2026. It's expected to house around 1.3 million Nvidia H100 GPUs.
- Hyperion: An even larger project planned for Louisiana, designed to scale up to a …
Archive Access
This article is older than 24 hours. Create a free account to access our 7-day archive.