News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Prometheus

Prometheus

Meta's first multi-gigawatt AI training cluster

Prometheus is a massive AI training cluster being built by Meta. It is the first of several multi-gigawatt facilities planned by the company. The cluster is scheduled to come online in 2026. It supports Meta's superintelligence research efforts.

Share on WhatsApp

  • First cluster comes online in 2026
  • Located in Ohio, United States
  • Part of multi-gigawatt infrastructure plan
  • Supports Meta Superintelligence Labs

What it is

Prometheus is an artificial intelligence training cluster developed by Meta. It represents the first phase of a broader infrastructure expansion involving several multi-gigawatt data centre clusters. The facility is designed to support advanced AI research and development. Meta states that this initial cluster will become operational in 2026. The project is part of the company's strategy to build elite compute resources for its superintelligence labs.

How it came about

The existence of the Prometheus cluster was first reported by the analyst firm SemiAnalysis in July 2025. The firm identified it as a one-gigawatt AI training cluster under construction in Ohio. Mark Zuckerberg later clarified the scope of the project via social media. He confirmed the name Prometheus for the first cluster and stated it would go online in 2026. He also announced plans for additional large-scale clusters, including one named Hyperion.

The compute

Prometheus is described as a multi-gigawatt AI training cluster. It is intended to provide industry-leading levels of computing power for researchers. Meta aims to offer the greatest amount of compute per researcher within its Superintelligence Labs. The specific hardware components, such as GPU types or networking architecture, are not detailed in the available information. The focus is on the sheer scale of power and capacity rather than specific chip models.

Where the GPUs are

The Prometheus cluster is being built in Ohio, United States. Meta plans to construct several similar multi-gigawatt facilities. The company illustrates the massive physical footprint of these future data centres by comparing them to the island of Manhattan. One such facility would cover a significant portion of that area. The Ohio site is the first of these titan clusters to be developed.

Customers and deals

The primary user of the Prometheus cluster is Meta itself. Specifically, it serves the Meta Superintelligence Labs. The infrastructure is designed to support internal research teams working on superintelligence. There is no indication that this compute capacity is offered as a commercial service to external customers or partners at this stage.

Money and ownership

Meta intends to invest hundreds of billions of dollars into compute infrastructure. This funding supports the construction of Prometheus and subsequent clusters like Hyperion. The company relies on capital generated from its existing business operations to finance these projects. This spending is part of a broader effort to advance AI capabilities, distinct from previous investments in virtual reality.

Its place in AI

Prometheus is central to Meta's superintelligence effort. It allows the company to concentrate talent and compute resources in one dedicated environment. The project highlights Meta's commitment to building large-scale infrastructure for AI training. It stands alongside other planned clusters such as Hyperion, which is designed to scale up to five gigawatts over time.

Questions people ask about Prometheus

When will the Prometheus cluster be ready?

Meta has stated that the Prometheus cluster is scheduled to come online in 2026. It is the first of several large-scale data centre clusters the company is currently building.

Where is the Prometheus cluster located?

The Prometheus cluster is being built in Ohio, United States. This location hosts one of the world's largest AI training clusters currently under construction by Meta.

What is the purpose of the Prometheus cluster?

Prometheus is designed to support Meta's superintelligence research. It provides high levels of compute power for researchers within the Meta Superintelligence Labs to advance AI frontiers.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Prometheus yet, so the facts on this page come from the sources listed above: theregister.com/2025/07/15/meta_datacenter_build_plan/, aimagazine.com/articles/mark-zuckerberg-reveals-100bn-meta-ai-superclu. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.

    It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

    Photo credits on this page

    The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

    The weekly digest

    Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.