Inside The Startup Launching AI Into Space
Inside The Startup Launching AI Into Space
Inside The Startup Launching AI Into Space
A new frontier in computing is unfolding above Earth, where the constraints of terrestrial infrastructure no longer apply. In this episode, we explore how a bold startup is redefining the limits of AI by launching powerful data centers into orbit—where sunlight never fades and cooling comes naturally from the void of space.
Starcloud has launched the first satellite equipped with an NVIDIA H100 GPU, marking a pivotal step toward orbital AI data centers. By harnessing uninterrupted solar power and using space’s cold vacuum for passive cooling, these centers eliminate the need for fresh water and reduce environmental impact. The team, backed by Y Combinator, built their 60-kilogram prototype in just 15 months and successfully launched it on November 2nd. Originally exploring space-based solar energy, they pivoted to computing as a more viable path amid falling launch costs. Their next mission, expected by October next year, will use NVIDIA’s Blackwell architecture to deliver ten times more computing power. As tech giants like Google, SpaceX, and Amazon begin eyeing orbital infrastructure, Starcloud aims to position space as the next critical layer of global AI compute—scalable, sustainable, and independent of Earth's strained resources.
00:00
00:00
Starcloud launched a satellite with an NVIDIA H100 GPU, the first of its kind in orbit
05:41
05:41
Orbital data centers have little market risk despite high technical challenges.
10:25
10:25
The first 60-kg satellite includes NVIDIA H100 GPUs for high-performance computing in space
10:25
10:25
Starcloud aims to prove modern data-center hardware can operate in orbit.
