Meta details Prometheus, the gigawatt scale Ohio supercluster where it trained Muse Wonder, citing over 85 percent pretraining goodput
We trained our Muse foundational models, including Muse Wonder, at Prometheus, our gigawatt-scale supercluster in New Albany, Ohio. Across a supercluster with hundreds of thousands of GPUs, even one underperforming host can slow an entire training run. We engineered Prometheus for resilience at scale, achieving 85%+ goodput on pretraining, 90%+ on reinforcement learning and ~47K GPU hours mean time between interruptions during Muse Wonder training. Take a look at the tech powering Prometheus:
Context
Keeping hundreds of thousands of GPUs productive is one of the hardest problems in frontier training, and Meta's figures give a rare look at how well one of the largest clusters runs.
The reliability figures are self reported by Meta. Prometheus is Meta's supercluster in New Albany, Ohio, and Meta released Muse Wonder on October 12, saying it was trained there.
- Meta: Muse Wonder, a frontier model for personal and professional AI agents
- Meta Data Centers: New Albany, where Ohio's community spirit meets the frontier of AI
- StorageReview: Meta Prometheus, inside the gigawatt supercluster scaling to 1 million GPUs
- Advance Intelligence: Mark Zuckerberg introduces Muse Wonder
Deep dive
On October 19, AI at Meta posted that it trained its Muse foundation models, including Muse Wonder, at Prometheus, its gigawatt scale supercluster in New Albany, Ohio.
Meta wrote that across a supercluster with hundreds of thousands of GPUs, even one underperforming host can slow an entire training run, so it engineered Prometheus for resilience at scale. It said the cluster achieved over 85 percent goodput on pretraining and over 90 percent on reinforcement learning, meaning the share of time spent on useful training rather than lost to failures and restarts, and a mean time between interruptions of about 47,000 GPU hours while training Muse Wonder. The post linked to a technical write up on the systems behind Prometheus.
The 24 second video opens on an aerial shot of the campus with the title Inside Prometheus, then moves through rows of server racks, cabling and networking hardware before ending on the Meta logo.
StorageReview, covering the write up, described Prometheus as scaling toward 1 million GPUs on Nvidia GB200 and GB300 NVL72 systems.
Meta's Muse Wonder announcement on October 12 said the model family was trained at Prometheus, covered in that day's deck.
What to expect next
Watch whether other labs publish comparable goodput and interruption figures, which would let outsiders compare how efficiently the largest clusters run. Meta's next test is whether Prometheus keeps these numbers as it scales toward a million GPUs, and whether the next Muse models arrive faster as a result. Also watch local debates over power use at gigawatt sites like New Albany, which Meta addresses in its own page on the community.