Nvidia H100 GPU Units By Company
This was originally posted on our Voronoi app. Download the app for free on iOS or Android and discover incredible data-driven charts from a variety of trusted sources.
As the demand for artificial intelligence skyrockets, companies across sectors are in a race to scale up their compute power, with billions being funneled into upgrading the infrastructure needed to support AI models.
Enter Nvidia’s H100 Tensor Core GPU, a top-of-the-line graphics processing unit designed specifically for artificial intelligence (AI) and high-performance computing (HPC) workloads.
This graphic visualizes the companies and organizations with the most Nvidia H100 GPUs using data from the 2024 edition of The State of AI Report, updated as of October 2024.
Meta Leads the Pack in H100 GPU Purchasing
Below, we show the number of Nvidia H100 GPUs various companies and organizations own, as of October 2024.
| Company/Entity | H100 GPU Count | Type |
|---|---|---|
| Meta | 350,000 | Private cloud |
| XAI/X | 100,000 | Private cloud |
| Tesla | 35,000 | Private cloud |
| Lambda | 30,000 | Public cloud |
| Google A3 | 26,000 | Public cloud |
| Oracle Cloud | 16,000 | Public cloud |
| Poolside | 10,000 | Public cloud |
| Magic | 8,000 | Public cloud |
| Andromeda | 3,632 | Private cloud |
| Scaleway | 1,016 | Private cloud |
| Hugging Face | 768 | Public cloud |
| DeepL | 544 | Private cloud |
| Recursion | 504 | Private cloud |
| Princeton | 300 | National HPC |
| Photoroom | 256 | Private cloud |
Companies like Meta and Tesla have been purchasing thousands of H100 GPUs to power their ambitious AI plans. While Nvidia does not officially disclose the price of its H100 products, estimates have priced it between $30,000 to $40,000 each for its 80GB version.
Nvidia’s high-performance H100 GPUs have significantly contributed to the company’s growing revenues, particularly in the data center segment.
In January 2024, Mark Zuckerberg announced Meta’s plan to acquire 350,000 Nvidia H100 GPUs by the end of the year, significantly boosting the company’s AI computing capabilities.
This massive investment in hardware is aimed at supporting Meta’s efforts to develop advanced AI models, including artificial general intelligence (AGI), a theoretical branch of artificial intelligence where AI achieves “human-level learning, perception and cognitive flexibility.”
During Tesla’s Q1 2024 earnings call, CEO Elon Musk revealed that Tesla had 35,000 H100 chips, which it plans to use to develop several of its AI projects, including the Robotaxi and a supercomputer called Dojo.
https://www.visualcapitalist.com/which-companies-own-the-most-nvidia-h100-gpus/
