SpaceX CEO Elon Musk has made a bold claim that his company’s Starship rocket could deliver 300 gigawatts (GW) of solar-powered artificial intelligence (AI) satellites into orbit every year. Musk’s announcement comes amid a growing interest in space-based data centers, a concept that could transform the way the world processes information. However, the billionaire entrepreneur also highlighted a major challenge that stands in the way of this ambitious plan.
Starship and the Promise of Orbiting AI Satellites
Musk shared his insights during a conversation that was later posted on social media by Salesforce CEO Marc Benioff. Benioff highlighted Musk’s point about the cost-effectiveness of operating data centers in space, noting that orbiting centers could use continuous solar energy without needing batteries, making them cheaper than traditional on-ground data centers.
In response, Musk emphasized the scale of what Starship could achieve. He said, “Starship should be able to deliver around 300 GW per year of solar-powered AI satellites to orbit, maybe 500 GW.” To put this into perspective, the average electricity consumption in the U.S. is about 500 GW. Musk suggested that with 300 GW per year, AI computing in space could surpass the processing power of the entire U.S. economy within just two years.
This would make orbiting data centers not only a futuristic idea but a serious contender for handling the world’s growing AI and computational needs. With Starship capable of moving large amounts of “tonnage to orbit,” the technical hurdle of getting massive solar AI satellites into space seems largely solved—at least in Musk’s vision.
The Major Challenge: Chip Production
Despite the promising possibilities, Musk pointed out a major “piece of the puzzle” that remains unsolved: chip production. AI satellites require advanced computer chips to function, and producing enough of these chips to meet demand is a massive challenge.
Musk referenced Tesla’s proposed Terafab, a large-scale chip manufacturing facility being developed in partnership with Intel. This “TeraFab” is designed to produce the chips necessary for both Tesla’s own AI projects and for space-based applications like Starship’s satellites. However, until large-scale chip production ramps up, delivering 300 GW of solar AI satellites per year will remain out of reach.
Elon Musk sparks frenzy by resurfacing Harris’ push to silence Trump on social media
Musk’s comments also come against the backdrop of Tesla shutting down its Dojo supercomputer team earlier this year. This decision underscores the complexity of AI infrastructure and how chip manufacturing is a critical bottleneck even for companies with ambitious space plans.
Space-Based Data Centers Are Gaining Interest
Musk’s announcement aligns with a wider movement among tech giants exploring space-based data centers. Last month, Blue Origin CEO Jeff Bezos discussed the potential for Gigawatt-scale data centers in orbit, claiming they could be more cost-effective than Earth-based operations. Musk, however, referred to Bezos’ idea with a touch of humor, calling him a “copycat.”
Alphabet CEO Sundar Pichai is also exploring the concept through Project Suncatcher, which aims to launch low Earth orbit data centers powered directly by solar energy. If successful, these projects could provide continuous, high-capacity AI computing without relying on terrestrial power grids, making global data processing faster and more sustainable.
The concept of space-based AI satellites is not just theoretical. With Starship’s ability to carry heavy payloads, these satellites could provide enormous computational power, support advanced AI applications, and potentially revolutionize industries from scientific research to autonomous driving.
SpaceX schedules successful Flight 11 of Starship to test final Version 2 rocket at Starbase
Musk’s announcement sheds light on both the potential and the current limitations of orbit-based technology. While Starship appears capable of launching hundreds of gigawatts of AI satellites each year, the world’s ability to produce the chips required for these systems remains a key obstacle.
As the space race for data centers heats up, companies like SpaceX, Blue Origin, and Alphabet are exploring ways to make orbiting AI computing a practical reality. However, the success of these ambitious projects will depend not just on rockets and satellites, but also on solving the technical challenges of mass chip production—the essential piece of the puzzle Musk identified.