Meta Prometheus: Inside the Gigawatt Supercluster Scaling to 1 Million GPUs on GB200 and GB300 NVL72
Meta published a detailed look inside Prometheus today, the gigawatt-scale AI supercluster it is building in New Albany, Ohio. The first phase is complete and has trained the Muse family of foundation models, including the Muse Wonder models, and Meta says the site is being scaled toward roughly 1 million GPUs. Two years ago, the The post Meta Prometheus: Inside the Gigawatt Supercluster Scalin…
Meta's Prometheus is a gigawatt-scale build in New Albany, Ohio, with its first phase complete and a stated path toward roughly 1 million GPUs on GB200 and GB300 NVL72. Deployments at that scale put sustained pressure on HBM supply, since every accelerator depends on it, and that tends to tighten allocation across the wider memory market. For enterprise buyers outside hyperscale, the practical risk is longer lead times and less availability on adjacent server memory like DDR5 RDIMM as suppliers prioritize the largest AI orders. If you're planning refreshes or capacity expansion this year, it's worth checking live inventory or requesting a quote before your timeline gets squeezed.