Starcloud


Starcloud, Inc., commonly known as Starcloud, is a United States–based company that designs, builds, and deploys data centers in space.

History

The company was founded in January 2024 in El Segundo, California under the name Lumen Orbit by Philip Johnston, Adi Oltean and Ezra Fielden. Soon after founding, in February 2024, the company relocated to Redmond, Washington to be close to the space and data center talent at Starlink, Amazon Leo, AWS, and Azure.
In Summer 2024, the company released a white paper and went through the Y Combinator startup accelerator in San Francisco, where they were selected by Tom Blomfield and raised one of the largest seed rounds ever at Y Combinator demo day.
In March 2025, the company rebranded to Starcloud after a legal challenge from Lumen Technologies and raised additional seed funding, bringing the total to approximately $34M. Investors include the scout funds of Sequoia and A16z, In-Q-Tel, NFX, Plug and Play, as well as angels, including AI expert Jan Leike.
In November 2025, Starcloud launched its first satellite, designated Starcloud-1, equipped with a Nvidia H100 GPU which Nvidia claimed was 100x more powerful GPU compute than had been in orbit before. The company described the mission as the first deployment of "data-center-class GPU compute" in orbit.
Starcloud stated that the project aimed to explore how orbital conditions, including continuous solar exposure and radiative cooling, could support large-scale computing in space.
In December 2025, Starcloud became the first company to operate a large language model on a high powered GPU onboard a spacecraft, and the first to perform in-orbit training of a large language model.
File:Starcloud-1 deploymet horizontal.png|thumb|400x706px|Starcloud-1 deployment from a SpaceX Falcon 9 rocket on November 2nd, 2025

Partnerships

In 2024, Starcloud was selected as a member of the Nvidia Inception program, the Google Cloud Accelerator, and the Defense Innovation Unit Accelerator.
In October 2025, Crusoe and Starcloud announced an agreement, as Crusoe will deploy its "Crusoe Cloud" platform on a Starcloud satellite planned for late 2026. Under this agreement, GPU capacity from orbit is expected to be offered from early 2027.
Starcloud-1 is also using the H100 to run high-powered inference on Synthetic-aperture radar data from Capella Space in order to draw insights on orbit for the first time, without needing to downlink the data before being able to analyse it.