Starcloud secures funding for orbital artificial intelligence data centers
Starcloud has raised a $250 million extension to its Series A funding round, bringing its valuation to $2.3 billion.
2 min read, Saturday, 22 August

Starcloud, a startup focused on deploying artificial intelligence inference capabilities in orbit, has expanded its March Series A funding round with an additional $250 million. This extension values the enterprise at $2.3 billion, backed by participants including Manhattan West Ventures, Nvidia, and Cisco.
Key facts
- Founder
- Philip
- Founder
- Ezra
- Founder
- Adi
Infrastructure and Manufacturing
The newly acquired capital will directly fund a larger manufacturing facility in Woodinville, Washington. Spanning 100,000 square feet, the site will support production lines for the startup's growing hardware ecosystem, currently managed by a team of 25 employees.
The primary focus remains the development of Starcloud-3, an orbital data center spacecraft designed to utilize SpaceX's forthcoming Starship vehicle. Prior to that deployment, the company plans to launch two 8 kW compute satellites, known as Starcloud-2, on rideshare missions in 2027 to serve customers such as United States government agencies.
Navigating Launch Constraints
Securing dependable transportation to orbit remains a primary logistical challenge. Leadership notes that market capacity is tightening, particularly with the scheduled phase-out of the Falcon 9 program in 2028 and limited availability among alternative launch providers.
While the long-term vision depends on Starship reducing transport costs to enable a competitive orbital inference layer, upcoming schedules face adjustments. SpaceX recently indicated delays for booster catch attempts and upcoming flight tests, making forward planning complex for satellite operators.
Hardware Development and Nvidia Partnership
Starcloud operates as the first entity to run a terrestrial Nvidia H100 data center graphics processing unit in orbit and train a model using it. Insights gathered from these operations are being shared with Nvidia to assist in the development of the Vera Rubin Space-1 chip, a specialized processor intended for space environments.
Engineers are currently addressing several specific design considerations for the upcoming space-ready hardware, expected to fly in late 2028:
- Thermal dynamics between operating temperatures and radiator sizing
- Placement strategies for radiation shielding
- Mechanical ruggedization to withstand launch vibrations
As the industry continues to evolve, the intersection of aerospace logistics and high-performance computing will define the next phase of infrastructure deployment beyond Earth's atmosphere.
Source: Original announcement
Worth a read?



Comments