
Building data centers in space offers a solution to the escalating energy and regulatory constraints facing terrestrial AI compute. By leveraging rapidly declining launch costs and the abundance of solar energy in orbit, Starcloud aims to deploy massive compute constellations that bypass Earth-bound infrastructure bottlenecks. Key engineering hurdles, such as thermal management in a vacuum and radiation exposure, are addressed through innovative immersion cooling and the use of high-performance, off-the-shelf components. Philip Johnston, CEO of Starcloud, emphasizes that booking launch capacity early acts as a critical forcing function for development. With plans to scale to 88,000 satellites, the company seeks to provide gigawatts of new compute capacity, positioning itself as a vital component of future national security and technological infrastructure.
Sign in to continue reading, translating and more.
Open full episode in Podwise