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’s 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’s NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Eos

Eos

Nvidia's DGX SuperPOD AI supercomputer for internal research

Eos is a large-scale AI supercomputer built and operated by Nvidia. It uses thousands of H100 GPUs to train advanced models and supports the company's own AI development efforts.

Share on WhatsApp

  • Ranked No. 9 on the TOP500 supercomputer list
  • Uses 4,608 Nvidia H100 Tensor Core GPUs
  • Delivers 18.4 exaflops of FP8 AI performance
  • Built on DGX SuperPOD architecture

What it is

Eos is a data-center-scale supercomputer owned and operated by Nvidia. It serves as an internal resource for the company’s developers to create AI breakthroughs. The system is built on the DGX SuperPOD architecture, which integrates accelerated computing hardware with optimized software. It is designed to handle massive workloads such as training large language models and running complex simulations. Eos ranked ninth on the TOP500 list of the world’s fastest supercomputers at its reveal.

How it came about

Nvidia revealed Eos in November 2023 during the Supercomputing 2023 trade show. The name comes from the Greek goddess who opens the gates of dawn each day. An editor’s note in February 2024 clarified that there are two distinct Eos systems. One is the primary supercomputer described here, while a separate sister system was used for MLPerf training benchmarks earlier that same month.

The compute

The primary Eos system contains 576 DGX H100 units, totaling 4,608 Nvidia H100 Tensor Core GPUs. It delivers 18.4 exaflops of FP8 AI performance. The cluster uses Nvidia Quantum-2 InfiniBand networking with In-Network Computing technology. This setup supports data transfer speeds up to 400Gb/s, ensuring low latency and high throughput for moving large datasets. Software tools include Nvidia Base Command and Nvidia AI Enterprise to manage the infrastructure.

Its place in AI

Eos acts as a showcase for Nvidia’s full-stack technology capabilities at scale. It supports the development of generative AI applications across fields like drug discovery and autonomous machines. By providing a purpose-built AI engine, it helps accelerate the creation of AI-infused applications. Its high ranking on global supercomputer lists highlights its computational power relative to other academic and government systems.

Questions people ask about Eos

What is the Eos supercomputer used for?

Eos is used by Nvidia developers to create AI breakthroughs. It handles large workloads like training large language models, recommender systems, and quantum simulations using accelerated computing infrastructure.

How many GPUs are in the Eos system?

The primary Eos supercomputer features a total of 4,608 Nvidia H100 Tensor Core GPUs. These are housed within 576 DGX H100 systems that make up the cluster.

What networking technology does Eos use?

Eos uses Nvidia Quantum-2 InfiniBand networking with In-Network Computing. This architecture supports data transfer speeds of up to 400Gb/s to facilitate rapid movement of large datasets.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Eos yet, so the facts on this page come from the sources listed above: blogs.nvidia.com/blog/eos/. 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.