1. Home
  2. / Science and Technology
  3. / Elon Musk surprises by building a colossal supercomputer in just 19 days with thousands of GPUs: Nvidia would take a year, and even its CEO recognized the incredible achievement!
reading time 3 min read Comments 0 comments

Elon Musk surprises by building a colossal supercomputer in just 19 days with thousands of GPUs: Nvidia would take a year, and even its CEO recognized the incredible achievement!

Written by Valdemar Medeiros
Published 06/11/2024 às 08:44
Elon Musk surprises by building a colossal supercomputer in just 19 days with thousands of GPUs: Nvidia would take a year, and even its CEO recognized the incredible achievement!
Photo: Playback/Youtube

Elon Musk's new $10 billion supercomputer was installed in just 19 days. Supercomputer with thousands of graphics cards promises to boost AI industry and oust others from the market.

A supercomputer with thousands of graphics cards has been rapidly built for xAI, a company founded by Elon Musk that focuses on developing artificial intelligence. The billionaire has positioned xAI as a direct competitor to OpenAI, the company behind ChatGPT. Elon Musk's supercomputer reportedly required an investment of nearly $10 billion, illustrating Musk's ambition to position himself as a key player in AI. Even the CEO of Nvidia was impressed, check out the details!

Understand the purpose of the supercomputer with thousands of video cards 

According to information provided by French portal JeuxVideo According to IGN, this Elon Musk supercomputer was designed to perform massive calculations linked to artificial intelligence, a growing need in the sector.

The project, carried out at lightning speed, not only highlights the billionaire's capacity for innovation, but also the current challenges in the race for computing power and artificial intelligence.

Even the CEO of Nvidia, Jensen Huang, expressed admiration for the speed with which the supercomputer with thousands of video cards was completed, explaining that this type of project normally takes four years.

According to Video games, the main difference here is that Musk and his team reduced that period to less than three weeks (19 days), an industry first. According to Huang, nothing like it has been done before.

Elon Musk's supercomputer project didn't just involve assembling components, but building an entire factory. This factory, which houses Nvidia H200 video cards, was equipped with advanced liquid cooling and power supply systems to manage the 100 GPUs required for this type of intensive computing.

Supercomputer uses the fastest video cards on the market

The H200 graphics card on Elon Musk's supercomputer is the fastest on the market, according to Nvidia. It is this model that is currently at the center of many projects related to artificial intelligence.

Nvidia GPUs have become a scarce resource due to the AI ​​boom. This has driven up prices and made it difficult to buy large quantities of these processors.

The most notable element of this supercomputer with thousands of graphics cards is not only the assembly speed, but also the implementation of the first phase of AI training on this supercomputer.

Going from the conceptual phase to functional use in such a short time is an impressive feat, especially since the Nvidia estimates that it takes an average of three years just to plan a similar data center.

The speed of the project has raised questions about how Musk and Nvidia managed to collaborate to achieve this feat. Huang noted that the final year of a project of this type is often spent installing and commissioning the equipment, a step that Musk's team was able to significantly reduce.

Musk's company becomes a player capable of competing with giants 

The coordination between xAI and Nvidia certainly played an essential role, although the exact details of this collaboration remain unclear.

This achievement is significant not only for what it represents in terms of technical capability, but also for what it suggests about the future of AI. The computational requirements for training and running artificial intelligence models are becoming increasingly gargantuan.

The ability to quickly build infrastructures tailored to these needs provides a major strategic advantage in a highly competitive industry. With Elon Musk’s supercomputer, xAI becomes a player capable of competing with giants such as OpenAI, Microsoft and Google. Accelerating the development of such infrastructures could also inspire other tech companies to rethink their construction and optimization processes to keep up.

Be the first to react!
React to article
Register
Notify
guest
0 Comments
Older
Last Most voted
Feedbacks
View all comments
Valdemar Medeiros

Journalist in training, specialist in creating content with a focus on SEO actions. Writes about the Automotive Industry, Renewable Energy and Science and Technology

Share across apps
0
We would love your opinion on this subject, comment!x