SpaceX and NVIDIA to Create AI Satellites

According to Ixbt.com, the outlet reports.

According to Interesting Engineering, each next-generation satellite will be equipped with NVIDIA Rubin graphics processors (GPUs) and NVIDIA Vera central processing units (CPUs). This is SpaceX’s official disclosure of the project’s hardware configuration and has attracted the attention of specialists.

Priority for Elon Musk and NVIDIA Technologies

Discussing the project’s technical foundation and future plans, SpaceX CEO Elon Musk said the company had decided to build all of its AI infrastructure exclusively on NVIDIA hardware. According to him, the company considers the Vera Rubin architecture the best solution currently available.

During the second-quarter earnings report, Elon Musk also said the company plans to increase the total computing power of its infrastructure to 10 GW by the end of 2027. However, this figure applies to the company’s entire computing infrastructure, not just its satellite constellation.

Testing and Production Timelines

According to the source, testing of the first Starmind AI1 prototype is scheduled for early 2027. If all work proceeds according to schedule, serial production is expected to begin later that same year.

Nevertheless, SpaceX has not yet disclosed how many satellites will be included in the system, what computing performance they will provide in orbit, or how the process of obtaining the necessary permits is progressing. The project’s success will largely depend on how technical challenges in the space environment are resolved.

The Specific Challenges of Orbital Computing Systems

Computing systems operating in space must function under intense radiation, dissipate heat efficiently in a vacuum, and have a stable power supply. They must also be able to recover from failures without regular maintenance and support high-speed data exchange.

SpaceX is still keeping the data storage architecture, network structure, and precise performance figures of future satellites confidential. However, it is clear that the selected NVIDIA platform provides the project with a robust hardware foundation.