Cowboy Space Raises $275M to Build Data Centers in Orbit

Cowboy Space has closed a $275 million funding round at a $2 billion valuation, bringing together investors Index Ventures and Blossom Capital. The company plans to do something unusual: turn the upper stages of rockets — the parts that normally separate and fall away — into data centers that stay in orbit and process information from Earth.

Founded in 2024 by Baiju Bhatt, co-founder of Robinhood Markets, Cowboy Space achieved a valuation of $1 billion in less than two years. The company originally started as Aetherflux to develop solar power systems in space, then rebranded and shifted its focus to orbital computing infrastructure.

How Orbital Data Centers Work

Rather than launching dedicated satellites designed specifically as computers, Cowboy Space's approach is to repurpose rocket upper stages themselves. After a rocket delivers its main payload to space, that upper stage — usually a large, powerful section built to push cargo into orbit — stays operational as a floating data center.

This matters because satellites designed specifically for computing face real constraints. They have limited power (from their solar panels), struggle to shed heat in the vacuum of space, and are hard to upgrade once launched. Rocket upper stages are bigger and more robust, with larger power systems and more room for server equipment and cooling hardware. Essentially, Cowboy Space is taking machines already designed to operate in the harshness of space and putting them to work processing data instead of being discarded.

Sitting in low Earth orbit — roughly 400 to 1,200 kilometers up — these data centers would communicate with ground stations with minimal delay, measured in fractions of a second rather than the full second or more that traditional satellites further out require. For applications that need real-time response, that speed difference is critical.

Why This Matters Now

Demand for computing power in space is growing. Companies that image Earth for agriculture, climate research, or surveillance need to process enormous amounts of data; shipping it all back to Earth wastes time and bandwidth. Autonomous systems operating in space need local computing resources so they can make decisions without waiting for instructions from the ground.

This pattern feels familiar if you know the history of cloud computing. In the late 1990s and early 2000s, companies like Amazon and Google realized they could sell access to spare computing capacity they were already building for themselves. That idea eventually became AWS and Google Cloud, reshaping how businesses operate. A similar shift appears to be starting in space: as launch becomes routine and cheaper, companies are asking what else they can do with expensive orbital assets.

SpaceX's Starlink constellation offers a parallel example. Originally designed to provide internet coverage to remote parts of Earth, it has evolved into a platform for other services, including direct connections to phones and government applications. Cowboy Space's move from building solar panels in space to building data centers follows the same logic — adapting a core capability (getting things into orbit and keeping them running) to wherever the market opportunity grows.

The Technical and Regulatory Reality Check

Operating data centers in orbit introduces challenges you would not face on Earth. Radiation at orbital altitudes can damage computer chips, so hardware must be hardened to tolerate constant exposure. Heat is harder to shed in the vacuum of space, requiring specialized cooling systems. Power comes from solar arrays, which means dealing with the regular darkness of eclipse periods — when the station passes through Earth's shadow — and maintaining battery backup.

Regulatory approval is still being sorted out. The FCC has rules for communications satellites, but no clear framework yet for persistent computing platforms that store and process data. Cowboy Space will need approval from the FAA for launches and from the FCC for transmitting data back to Earth.

Orbital mechanics add another layer. Depending on how high and at what angle Cowboy Space places these data centers, they will only have clear line-of-sight to ground stations for limited windows of time as they orbit. Providing continuous service means either deploying many stations or coordinating with other satellite networks to maintain coverage.

What This $275 Million Needs to Do

This funding level shows both the scale of space ventures and investor faith in the orbital computing idea. Launching even reusable rockets costs millions per flight, and building reliable systems that operate in orbit for years is expensive in ways that Earth-based cloud computing is not. Software companies can iterate quickly and affordably; space infrastructure requires large upfront spending before it generates any revenue.

The funding will likely support demonstration missions — proof that the concept actually works — and development of standardized modules that can turn different rocket upper stages into data centers. The inclusion of Blossom Capital, a European investor, may point toward expanding beyond the United States market or managing regulatory approval in multiple countries.

Broader Implications

The bigger picture here involves the maturation of the space economy beyond just launching things and running satellite internet. As getting to orbit becomes more routine and cheaper, companies are asking what high-value work they can do once they are there. Orbital data centers offer something terrestrial cloud providers cannot: geographic redundancy so extreme that no ground-based attack or disaster can compromise your computing.

Governments and defense organizations may see particular value in computing infrastructure that sits above the Earth, beyond the reach of conventional threats. If Cowboy Space succeeds in building reliable orbital infrastructure, it will likely spark similar ventures. Much like the early days of cloud computing, whichever companies establish themselves as dependable providers of orbital computing may capture a significant share of a new market as it grows.