Editorial: Orbital data centers are a crazy idea. Go for it

The Editors, Bloomberg Opinion on Aug 1, 2026

By clicking submit, I authorize Arcamax and its affiliates to: (1) use, sell, and share my information for marketing purposes, including cross-context behavioral advertising, as described in our Privacy Policy , (2) add to information that I provide with other information like interests inferred from web page views, or data lawfully obtained from data brokers, such as past purchase or location data, or publicly available data, (3) contact me or enable others to contact me by email or other means with offers for different types of goods and services, and (4) retain my information while I am engaging with marketing messages that I receive and for a reasonable amount of time thereafter. I understand I can opt out at any time through an email that I receive, or by clicking here

Why not just build them in space?

As data centers face rising costs, strained resources and growing political opposition, a number of entrepreneurs — Jeff Bezos, Elon Musk, Eric Schmidt — have raised the possibility of processing and storing data aboard satellites in low-Earth orbit, using large solar arrays for power. It’s a crazy idea, and well worth supporting.

Data centers are an essential element of the digital economy, powering everything from cloud computing to artificial intelligence. Problem is, they also require a lot of land, energy and water. Building them often entails navigating a morass of permitting rules, litigation, and occasional protests, and may cost billions. Some politicians are calling for moratoriums or even an outright ban.

Moving the process into space could (in theory) elide such problems. To an extent, it’s already happening. Last year, Starcloud Inc. sent an Nvidia Corp. processing unit into orbit. Axiom Space Inc. launched two data-center nodes in January, which it says “will lay the foundation for space-based cloud computing.” Blue Origin LLC and SpaceX are both planning huge constellations of data-processing units. Industry proposals envision putting hundreds of thousands of satellites into orbit over the next few decades.

Practical impediments aside, this could have many benefits. Space-based data centers won’t need to overcome zoning hurdles, secure rights-of-way or wait for utility interconnections. Modular designs may make them easier to scale. In the vacuum of space, they won’t consume water, generate noise, or agitate municipal governments. Nor will anyone complain about how ugly they are.

Energy could be an even bigger benefit. Data centers are expected to fuel almost half the growth in US electricity demand by 2030, worsening bottlenecks and straining the grid. In space, by contrast, energy is plentiful. In favorable orbits, satellites can access near-continuous high-intensity solar power while avoiding weather and atmospheric attenuation. As a result, a given array in space might generate five times more energy than its terrestrial equivalent, according to one industry analysis.

More intriguingly, orbital data centers — which will require repeated deliveries of hardware, arrays and so on — could create sustained demand for heavy-lift rocket launches. That could spur process innovation, mass production, and standardization, thereby slashing costs. One study by Google researchers found that (to simplify a bit) 180 launches of SpaceX’s Starship rocket per year, at the company’s estimated learning rate, might reduce launch prices to as low as $200 per kilogram over a decade; that would not only help make space-based data centers cost-competitive but potentially open up vast new opportunities in orbit, both scientific and commercial.

All that said, pitfalls abound. Cooling these systems will require radiators to disperse heat into space, an unproven process at such scale. Radiation may degrade chips or create data disruptions. Hundreds of satellites may be needed to match the output of a single big data center — plus more for servicing and repairs — potentially creating congestion, collisions and debris. By any reckoning, launch economics will be a challenge.

In short, orbital data centers may prove to be a completely irrational business venture. That’s no reason to oppose them. Instead, policymakers should focus on building the infrastructure to support space launch more broadly — steps that would be productive even if orbital data centers don’t pan out, and all the more so if they do.

As one example, federal agencies should work with orbital-processing companies on tasks such as monitoring space debris or filtering Earth-observation imagery. If in-space processing can reduce transmission times and speed analysis, both public and private sectors will gain. Similarly, a concerted effort to expand spaceport capacity, reduce red tape for launches, and fund R&D on relevant technology — such as lightweight radiators, optical laser links or robotic assembly — could have immense benefits for the broader space economy, not to mention terrestrial science.

Whatever the fate of data centers, such steps will only accelerate American spacefaring and innovation. There’s little to lose and a universe to gain.

____

The Editorial Board publishes the views of the editors across a range of national and global affairs.