Menu

Search

  |   Technology

Menu

  |   Technology

Search

Elon Musk’s xAI Installs 100,000 Nvidia H200 GPUs in 19 Days—Jensen Expected It to Take 4 Years

Elon Musk’s xAI challenges traditional GPU setup timelines, finishing in just 19 days. Credit: EconoTimes

Elon Musk’s xAI team accomplished the rapid installation of 100,000 Nvidia H200 GPUs in only 19 days, a timeline that Nvidia CEO Jensen Huang says normally stretches to four years in data center projects.

Musk's xAI Defies GPU Setup Timelines

In an astounding 19 days, Elon Musk and the xAI team accomplished an engineering feat by establishing a supercluster of 100,000 H200 Blackwell GPUs. On X, Jensen Huang, CEO of Nvidia, shared with Tesla Owners Silicon Valley a tale of Elon Musk's remarkable installation skills.

As awe-inspiring and respectful, Huang characterizes Musk's 19-day adventure as "superhuman." Reportedly, the xAI team completed the transition from "concept" to "gear" compatible with Nvidia in about three weeks. This also involves training the first AI on the freshly constructed supercluster using xAI.

Building the Supercluster: From Factory to Functionality

The entire procedure began with the construction of the enormous X factory, which would house the GPUs. It continued with the installation of electricity and liquid cooling systems across the entire facility, ensuring that all 200,000 GPUs could function properly.

Per Toms Hardware, getting the hardware and infrastructure supplied and installed in a precise and coordinated manner required a great deal of cooperation between the engineering teams at Nvidia and Elon Musk.

Jensen Praises Musk's Speed and Precision

According to Huang, a typical data center would need four years to do what Elon Musk and his team accomplished in just nineteen days. The first three years would be devoted to planning, while the last year would be occupied with shipping, installing, and getting everything up and running.

Additionally, Huang describes in great detail the intricate nature of the networking on Nvidia's hardware. Networking Nvidia's gear isn't the same as networking regular servers in a data center, he says. "The number of wires that goes in one node...the back of a computer is all wires."

Four Years vs. 19 Days: How xAI Beat the Odds

No other firm is likely to replicate Elon Musk's integration of 100,000 H200 GPUs for quite some time, given it has "never been done before" (according to Jensen Huang).

  • Market Data
Close

Welcome to EconoTimes

Sign up for daily updates for the most important
stories unfolding in the global economy.