Musk shares video of Huang delivering the first AI-optimised GPU to OpenAI in 2016. Watch

Elon Musk appeared visibly proud in the video, beaming at the silvery box as Nvidia CEO Jensen Huang explained its components. (Image credit: @cb_doge/X)

Elon Musk on Monday shared a video from 2016 that captures a pivotal moment in the history of artificial intelligence — Nvidia CEO Jensen Huang personally delivering the company’s first AI-optimised GPU server to OpenAI. At the time, OpenAI was a non-profit co-founded by Musk and Sam Altman.

The clip, posted on X, shows Huang arriving at OpenAI’s San Francisco office with the DGX-1, a $129,000 supercomputer designed specifically for AI workloads. The device, described by Nvidia as “the world’s first supercomputer dedicated to artificial intelligence,” contained eight interconnected GPUs and could process in two hours what traditional CPUs would take six days to compute, Bloomberg reported in 2016.

Musk, who had long warned about the dangers of unchecked AI, appeared visibly proud in the video, beaming at the silvery box as Huang explained its components.

Jensen delivering the first AI-optimized GPU to OpenAI in 2016 https://t.co/o4Qx930PDX

— Elon Musk (@elonmusk) August 11, 2025

The DGX-1,  Huang’s visit came just days before Intel’s annual developer forum — a strategic move that underscored Nvidia’s ambition to dominate the AI hardware space and the moment marked the beginning of a deep collaboration between Nvidia and OpenAI, one that helped accelerate the development of large language models like ChatGPT.

Years later, Musk would leave OpenAI’s board citing potential conflicts with Tesla’s growing AI focus. In 2023, he launched xAI, a direct competitor, and later sued OpenAI over its shift to a for-profit model. The lawsuit had sought a jury trial and for the company to pay back any profit they received from the business. But it was withdrawn after OpenAI pushed back against Musk’s claims, calling them “incoherent” and “frivolous” and published a blog post that included several of Musk’s emails from OpenAI’s initial days that appeared to show Musk acknowledging the need for the company to make a lot of money to fund the computing resources needed to power its AI ambitions.