A five-month-old company with six employees just filed with the FCC to launch 100,000 satellites.
Not communication satellites, not Earth observation satellites, but data centers in orbit. The space hiring it triggers touches a talent pool the sector has never tapped before.
Orbital, based in Los Angeles, wants to put 10 gigawatts of computing power in space – roughly the same amount of new electricity capacity the entire US power grid added last year. The satellites would be 100-kilowatt class, sitting in low Earth orbit between 500 and 850 kilometers, with solar arrays and radiators spanning about 100 meters each.
The plan is early, and Orbital has $5 million in pre-seed funding, a team of six people from SpaceX, Amazon, and Northrop Grumman, and a demo mission planned for next year with a single GPU. The first real compute satellite isn’t expected until 2028. The full constellation is into the next decade.
But Orbital isn’t the only story. It’s the latest signal in a pattern that’s been building all year – and that pattern is about to change who space companies are hiring in 2026 and beyond.
Who’s Building Orbital Data Centers in 2026
Orbital’s filing landed three weeks after the company came out of stealth. It joins a growing list of companies betting that the future of AI computing isn’t in a warehouse in Virginia – it’s in orbit.
SpaceX filed with the FCC in January for up to one million orbital data center satellites. That filing came days after the SpaceX-xAI merger and fits into a broader strategy that includes the $1.75 trillion IPO. SpaceX’s S-1 showed its AI segment burning $7.7 billion in capital expenditure in Q1 alone, and the company has said it plans to deploy data centers in space as early as 2028.
Starcloud has proposed an 88,000-satellite constellation with 200-kilowatt-class spacecraft. Cowboy Space has filed its own orbital compute plans. Even Vast – the station company that recently expanded into satellite buses – is offering an optional NVIDIA AI compute module on its platform.
The thesis behind all of them is the same: AI is eating more power than the ground can provide. Data centers need electricity, cooling, and land, and all three are running short. Space has constant sunlight for power, the vacuum of space for cooling, and no neighbors to complain about noise or water usage.
Whether 100,000 or a million orbital data centers actually get built is a question for the next decade. But the investment is moving now, and the AI jobs in space it creates are already being filled.
The New Engineering Roles Orbital Data Centers Are Creating
Here’s why this matters for anyone thinking about where space careers are heading.
Every other trend we’ve covered this year – Artemis, Golden Dome, the Space Force budget, commercial constellations – creates demand for traditional space engineers. Systems engineers, GNC specialists, propulsion engineers, flight software developers – the people who’ve always built spacecraft.
Orbital data centers are different. They need those people too – someone has to build the satellite bus, design the power systems, manage the thermal environment in orbit. But they also need an entirely new category of talent that the space sector has never recruited before.
Data center architects who understand how to design computing infrastructure at scale.
The people who’ve spent their careers at AWS, Google, Microsoft, and Meta designing the cooling systems, power distribution, and rack layouts for terrestrial server farms. Their knowledge of thermal management, power efficiency, and high-density computing translates directly to the orbital problem – the physics are different, but the engineering principles are the same.
GPU and chip-level engineers who understand how processors perform under extreme conditions.
Space adds radiation, vacuum, and thermal cycling to the engineering challenge. Nvidia’s involvement (through partnerships with Vast and others) signals that the GPU expertise currently concentrated in Silicon Valley is about to become relevant to space for the first time.
Optical networking specialists who can build the communication links between satellites and between orbit and the ground.
Orbital’s plan – like SpaceX’s and Starcloud’s – relies on optical inter-satellite links to move data between the compute nodes and back to Earth. The engineers who build these links are currently working in telecom and fiber optics. Space needs them.
Power systems engineers at a scale the space sector hasn’t seen.
A 100-kilowatt satellite is roughly ten times the power of a typical commercial communications satellite. Designing, deploying, and managing solar arrays and power distribution at that level is a different engineering problem than what most spacecraft power engineers have worked on.
This is a discipline convergence. Two talent pools – space hardware and terrestrial computing infrastructure -that have never overlapped are about to merge. The engineers at hyperscalers who’ve never considered a space career are suddenly relevant. And the space engineers who’ve never thought about data center architecture are about to need that vocabulary. For software engineers wondering how to get into the space industry, this may be the most accessible entry point yet.
Orbital Data Center Timeline: What’s Funded and What’s Still Speculative
Let’s be clear about timelines. Orbital has six people and $5 million. Their first GPU demo is next year, and full-scale deployment is years away. Even SpaceX’s million-satellite filing is aspirational at this point – the technology for orbital computing at that scale doesn’t fully exist yet.
But the hiring doesn’t wait for the technology to be ready. SpaceX is already building toward orbital compute through its xAI integration. Vast is offering AI modules on its satellite bus. The defense sector is investing billions in space-based data processing through programs like Golden Dome and the Space Force’s AI command-and-control experiments.
The companies that will lead orbital computing in 2030 are hiring the founding teams now. And those teams are being built from a talent pool that didn’t exist as a category twelve months ago.
What This Means for Space Careers
The space sector has always hired rocket scientists, satellite engineers, and mission operators. That’s not changing. But the orbital data center race is adding a new layer of demand – for engineers who understand computing infrastructure, thermal management at scale, GPU performance, and optical networking.
For the first time, a career at Google’s data center division or Meta’s infrastructure team is directly relevant experience for a space company. That’s a shift the sector hasn’t seen before, and it’s going to reshape who space companies recruit, where they recruit from, and what a “space career” looks like for the next generation of engineers.
The newest job in space isn’t building rockets. It’s building the servers that fly on them.