Starcloud Raises $250M for Orbital AI Data Center Expansion
Nvidia and Cisco back a $420M Series A extension to secure launch capacity for Starship-ready compute satellites.
Orbital AI infrastructure startup Starcloud has raised a $250 million Series A extension at a $2.3 billion valuation, led by Manhattan West Ventures with a $25 million direct strategic investment from Nvidia. The landmark round addresses severe terrestrial power bottlenecks by deploying high-density AI inference clusters into Low Earth Orbit. By securing early launch manifests on SpaceX’s upcoming Starship rocket, Starcloud aims to shield enterprise AI workloads from rising energy costs on Earth while providing low-latency space compute for commercial and defense operators.
Key Details
Starcloud confirmed the $250 million extension brings its total Series A funding to $420 million. In addition to Manhattan West Ventures and Nvidia, major investors participating in the round include Cisco, Benchmark, EQT, Soma Capital, NFX, 776, Cedar Capital, Goanna Capital, and Standard Capital. The capital will fund a new 100,000-square-foot satellite manufacturing facility in Woodinville, Washington, near existing space hardware hubs for SpaceX Starlink and Amazon Project Kuiper.
The primary driver for the capital raise is securing orbital launch capacity as the space transportation market tightens ahead of SpaceX phasing out its Falcon 9 vehicle in 2028. Starcloud has already requested FCC permission to operate up to 88,000 compute satellites in orbit. In 2027, Starcloud plans to deploy two 8 kW "Starcloud-2" compute satellites on rideshare flights to execute orbital inference tasks for U.S. government agencies, while preparing its massive "Starcloud-3" platforms for dedicated Starship payloads.
What This Means
As terrestrial AI data centers face severe power constraints, land-use battles, and environmental opposition over water usage, orbital compute is shifting from speculative science fiction to viable infrastructure. Operating GPUs in Low Earth Orbit offers continuous solar power generation without reliance on ground electricity grids or fossil fuel power plants.
However, moving AI workloads into space substitutes terrestrial power bottlenecks with orbital launch constraints. Starcloud’s massive funding round signals that frontier AI companies must now compete directly for rocket launch manifests to guarantee compute expansion. By partnering directly with chipmakers and satellite manufacturers, Starcloud is establishing an early moat in space-based artificial intelligence.
Technical Breakdown
Starcloud’s operational data from orbit is directly shaping next-generation AI hardware for harsh space environments:
- Orbital H100 Validation: Starcloud is currently operating the only active Nvidia H100 GPU in space on its Starcloud-1 testbed, successfully training models in orbit to establish thermal and radiation baseline metrics.
- Custom Silicon Collaboration: Data gathered from Starcloud-1 is being shared directly with Nvidia engineers to co-develop the Vera Rubin Space-1 chip, Nvidia’s first GPU built specifically for space environment survival.
- Thermal and Radiation Design: Key engineering priorities for Starcloud’s Woodinville facility include designing large-surface radiative cooling panels for vacuum heat dissipation and ruggedized radiation shielding to prevent bit-flips during launch vibration and cosmic ray exposure.
Industry Impact
For AI developers and enterprise cloud users, Starcloud’s expansion opens up a third compute tier alongside hyper-scaler data centers and edge devices. Defense and government agencies stand to gain immediate benefits through zero-latency, on-orbit processing of synthetic aperture radar (SAR) and high-resolution Earth observation data, eliminating the bandwidth bottleneck of downlinking massive raw datasets to ground stations.
For the semiconductor and space industries, Nvidia’s $25 million investment marks the first time a major GPU manufacturer has taken a direct equity position in an orbital compute startup. This strategic alignment ensures custom space-grade GPUs will enter mass production alongside terrestrial datacenter silicon.
Looking Ahead
Starcloud’s immediate priority is scaling production at its Washington facility ahead of the 2027 launch of Starcloud-2. The company’s long-term unit economics depend heavily on SpaceX demonstrating rapid reuse and cost reductions for Starship, enabling the deployment of Starcloud-3 orbital data halls by 2028.
As ground-based data centers face increasing regulatory scrutiny and energy caps, the race to build off-world compute infrastructure is accelerating. Success for Starcloud could establish space as the standard tier for high-throughput AI inference in the decade ahead.
Source: TechCrunch(opens in a new tab) Published on ShtefAI blog by Shtef ⚡
