Skip to content
Arne Noori
Aboutarne.ai
← All artifacts
Nvidia DGX-1

Nvidia DGX-1

Image creditFritzchens Fritz, via Wikimedia Commons (the GP100 chip of the Tesla P100 GPUs in the DGX-1, not the full server)
CC0 ↗

Nvidia DGX-1

Year
2016
Creator
Jensen Huang (CEO) and Nvidia engineering teams
Company
Nvidia
Location
Santa Clara, California; first unit delivered to OpenAI in San Francisco

Description

The DGX-1 is a server that Nvidia built for deep learning, with eight Tesla P100 graphics processors and two Intel Xeon processors. Nvidia introduced it in April 2016 at a price of $129,000.

Importance

The DGX-1 was a complete AI computer from Nvidia, and not only a graphics card. In August 2016, Jensen Huang delivered the first DGX-1 to OpenAI in San Francisco. In 2022, Nvidia announced its fourth generation of DGX systems, built around the H100, which is also in this room.

Interesting fact

Jensen Huang wrote a short note to Elon Musk and the OpenAI team on the case of the machine before he gave it to them.

Sources

  1. NVIDIA's insane DGX-1 is a computer tailor-made for deep learning ↗Engadget · 2016-04-05
  2. Nvidia Just Gave A Supercomputer to Elon Musk-backed OpenAI ↗Fortune · 2016-08-15
  3. NVIDIA Delivers DGX-1 Supercomputer in a Box to OpenAI ↗TOP500 · 2016-08
  4. NVIDIA Announces Hopper Architecture, the Next Generation of Accelerated Computing ↗NVIDIA Newsroom · 2022-03-22