Elon Musk, CEO of SpaceX, announced on August 24 (local time) that the company will launch a satellite equipped with artificial intelligence (AI) chips in the fourth quarter of next year.

Reuters Yonhap News

Reuters Yonhap News

View original image

He stated on social media platform X (formerly Twitter), "SpaceX has worked with Nvidia to design a Vera Rubin (Vera Rubin) NVL72 system optimized for the space environment," adding, "We plan to launch this into orbit in the fourth quarter of next year and significantly scale up by 2028." Musk emphasized that the data centers in satellite form are "much simpler, less expensive, denser, and lighter than existing racks."


The Vera Rubin NVL72 is Nvidia’s next-generation AI computing platform. It is a high-performance AI computation system that connects 72 Rubin GPUs with 36 Vera CPUs into a single unit. Earlier this month, SpaceX decided to use only Nvidia technology for its AI infrastructure in space. While Musk initially mentioned that the company would consider various designs utilizing multiple types of AI chips, he ultimately concluded that Nvidia’s system is best suited for AI infrastructure in space.

Musk: SpaceX to Launch AI-Powered Satellite in Q4 2027 View original image

When SpaceX was preparing for its initial public offering (IPO), it cited 2028 at the earliest as the timeline for the first launch of an orbital data center. However, in SpaceX’s first earnings release post-IPO, Musk updated this forecast to "next year," and now the company has specified the timing in more detail.



SpaceX is pushing ahead with the construction of a large-scale space data center network. In January, it applied to the U.S. Federal Communications Commission (FCC) for approval to build an AI computation network comprised of up to one million satellites. On June 8 of this year, the company also detailed its plan for an orbital AI data center powered by solar energy, in a video presentation. In this system, satellites would be powered by solar energy and directly perform AI computations while in orbit. Musk has previously predicted that within the next two to three years, space will become the most cost-effective location for running AI computations.


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing