This isn’t just about launching a few more satellites. SpaceX has filed an application with the US Federal Communications Commission for a network of a million satellites dedicated to creating an AI-powered data processing web. The scale is staggering, dwarfing the existing Starlink constellation of roughly 10,000 satellites. So, is this the next logical step for cloud computing, or a celestial land grab with concerning consequences?
What Exactly Is Orbital Computing?
Let’s get one thing straight: orbital computing is more than just sticking a server on a rocket and hoping for the best. It’s about creating a distributed computing network in space, where data is processed closer to where it’s collected—primarily from Earth-observation or communication satellites. Instead of beaming terabytes of raw data down to Earth, processing it, and then sending commands back up, the computation happens right there, in low Earth orbit.
Think of it this way. A terrestrial data centre is like a massive central library. If you’re a satellite (the reader) in orbit, you have to send a messenger all the way back to the library on Earth to look something up, wait for the answer, and then get the information sent back. Orbital computing puts a well-stocked, interconnected series of local library branches right in your neighbourhood, slashing the travel time. This is crucial for applications that need near-instantaneous processing, like climate monitoring, disaster response, or even future military surveillance.
The key argument, as reported by outlets like the BBC, is efficiency. By moving the processing offshore—or in this case, off-planet—proponents claim they can sidestep the massive energy and cooling demands of Earth-based data centres. But accomplishing this requires overcoming some serious technical hurdles.
The Celestial Tech Stack
Putting a computer in your office is easy. Putting one in the vacuum of space, where it’s simultaneously baked by the sun and frozen in shadow while being bombarded with radiation, is another matter entirely.
– Radiation-Hardened Hardware: Space is flooded with high-energy particles from the sun and cosmic rays. These particles can wreak havoc on standard electronics, causing data corruption or complete hardware failure. This is why missions rely on radiation-hardened hardware—specialised chips and components designed to withstand this hostile environment. They are more expensive and often less powerful than their terrestrial counterparts, creating a constant trade-off between resilience and performance.
– Inter-Satellite Networking: For a million satellites to act as one cohesive data centre, they need to talk to each other. This is where inter-satellite networking comes in. Using laser links, the satellites can pass huge amounts of data between themselves, creating a mesh network in the sky. This allows a job that starts on one satellite to be passed to another with spare capacity, all without ever touching the ground. It’s a foundational technology for making orbital computing a truly distributed system.
– Thermal Management: On Earth, we cool data centres with air and water. In the vacuum of space, you don’t have that luxury. Thermal management becomes a critical design challenge. Engineers must use passive methods like radiators to dissipate heat into space and reflective coatings to avoid absorbing too much solar energy. Keeping the electronics within their optimal temperature range is a constant, delicate balancing act.
The Economics of Launching Your Data
For decades, the cost of launching anything into space was astronomical. This is where the concept of launch economics has been completely rewritten by companies like SpaceX. With reusable rockets, the cost per kilogram to orbit has plummeted, making the idea of launching entire data centres at least financially plausible.
The business case hinges on a simple, yet unproven, calculation: is the cost of developing, launching, and maintaining a million satellites in orbit cheaper than the long-term energy and infrastructure costs of equivalent data centres on Earth? Musk’s camp argues that space-based solar power makes orbital data centres more sustainable. However, that claim conveniently ignores the enormous carbon footprint of manufacturing and launching a million satellites in the first place.
Then there’s the growing problem of space congestion. Musk has tried to downplay this, stating, “The satellites will actually be so far apart that it will be hard to see from one to another. Space is so vast as to be beyond comprehension.” While technically true, that vastness is becoming increasingly crowded, particularly in the useful low Earth orbit altitudes between 500 and 2,000 kilometres. The risk of collisions, creating more space debris in a cascading effect known as the Kessler syndrome, is a very real concern for everyone operating in space.
The SpaceX Gambit and the Astronomer’s Dilemma
The SpaceX proposal is nothing if not ambitious. The plan for a million satellites aims to serve “billions of users globally,” suggesting a scale that goes far beyond niche scientific or military applications. It hints at a future where our primary connection to the digital world is mediated through a network in the stars.
Yet, this vision is already causing conflict. In 2024, astronomers complained that radio waves from the existing Starlink network were already “blinding” their telescopes, interfering with humanity’s ability to observe the universe. A million more satellites, each a potential source of radio noise and light pollution, would only compound this problem. This sets up a fundamental clash between two different modes of exploring the final frontier: one looking out into the cosmos, the other using near-space as a platform for Earth-bound services.
Musk’s proposal also references a concept from the 1960s astronomer Nikolai Kardashev, hinting that this level of energy capture could move humanity towards a “Kardashev II civilisation”—a society capable of harnessing the entire energy output of its star. It’s a grand, almost mythic, justification. Is it genuine aspiration, or just savvy marketing to frame a commercial venture in world-historical terms?
Where Does This Go From Here?
If orbital computing can overcome its immense technical and economic challenges, its potential is undeniable. Near real-time global monitoring, ultra-low latency communications, and a less energy-intensive backbone for AI are all on the table. It could unlock new scientific discoveries and commercial opportunities we can barely imagine today.
However, the path is fraught with risk. We are treating low Earth orbit like an unregulated frontier, a new Wild West. Without robust international agreements on space traffic management, debris mitigation, and spectrum allocation, we risk cluttering our orbital highways to the point of unusability.
The dream of a clean, efficient orbital data centre must be weighed against the reality of launch pollution, space debris, and the potential blinding of our astronomical eyes on the universe. Innovation is vital, but so is stewardship. As we reach for the stars, we need to be careful not to trample them in the process.
What do you think? Is orbital computing a necessary evolution for our digital age, or are we creating an unsolvable problem for future generations? Let me know your thoughts below.
