Meta Details Ohio Prometheus Campus as First Phase Goes Live

Meta has provided an update on its Prometheus data center campus in New Albany, Ohio, a site the company plans to scale to one gigawatt. The property originally operated as a conventional Facebook data center. Meta says it has since converted the facility into the first phase of an AI-focused supercluster, and that this phase has already been used to train its Muse foundation models.

The campus currently houses hundreds of thousands of GPUs across several interconnected buildings, including traditional and AI-optimized data center halls as well as rapid-deployment structures. At full scale, Meta expects roughly one million distributed GPUs. The hardware includes Nvidia GB200 and GB300 systems, with next-generation Vera Rubin chips described as arriving soon.

Meta's engineers describe a networking design in which building-level Ethernet fabrics connect through a backend aggregation network spanning roughly 2 km to more than 100 km. The company also says the GB300 platform can use either its modular air-assisted liquid cooling or facility liquid cooling.

For developers and financiers, the takeaway is that hyperscale AI capacity is being created by converting existing data center sites, so power, cooling and networking upgrades will likely be central to feasibility and financing for similar projects.