Orbital Compute Baseline
The case for space datacenters
We have moved past the “is this possible?” phase of orbital datacenters.
We now proceed to “what is permissible” phase.
While terrestrial datacenters are sweating over power purchase agreements and grid constraints, a new consensus is forming 62 miles over our heads beyond the Kármán line. The narrative has shifted from pure science fiction to a necessity driven by the brutal math of AI energy consumption.
The loudest signal so far is Aetherflux putting a due date on the calendar for its first commercial orbital datacenter node in Q1 2027. Start the countdown.
The “Fudge Factor” in play will seek to bypass the lead times for terrestrial utility connections and deploy reliable, solar-dense compute in orbit. Demonstrating power beaming is a bold bet that disposable, high-cadence satellite launches can outpace the sluggish timelines of Earth-based infrastructure.
If you watched Gavin Baker articulate the case for space, the first principles are clear:
- Solar is ~30% more intense than on Earth, with up to 6x the effective irradiance due to 24/7 exposure
- You don’t need massive battery arrays if the sun never sets
- Radiators facing deep space offer a heat sink near absolute zero (effectively free cooling)
- Light travels roughly 30% faster in the vacuum of space than in fiber optics
Meanwhile, academia is providing the rigorous framework to back these claims, with recent papers in Nature Electronics validating the thermal and energy models of space-based edge computing. So, the solar race is on as well as the race for launch capacity (i.e. SpaceX Starship and Blue Origin New Glenn for heavy lift), we have the thermal modeling, and now, we have the launch dates.
Keep an eye on the launch cost in terms of cost per unit of weight. As that metric drops, things get even more interesting. Along the way, keep an eye on these market players:
| Company | Category | Focus | Key Partners |
|---|---|---|---|
| Aetherflux | Startup | Integrated Power & Compute (“Galactic Brain” constellation) | Apex Space, SpaceX |
| Amazon | Infrastructure | Orbital Connectivity Capability (Leo edge processing) | US Gov |
| Axiom Space | Infrastructure | Hosting (Server racks on commercial space station) | Voyager, AWS |
| Infrastructure | Solar AI Constellation (Project Suncatcher) | Planet Labs | |
| HPE | Enterprise | Edge Computing on ISS (Spaceborne Computer-2) | NASA, OrbitsEdge |
| KP Labs | Software / Hardware | Smart Mission Ecosystem (AI Data Processing Units) | ESA, AAC Clyde Space |
| Lonestar | Startup | Lunar Storage & DR (Disaster Recovery on the Moon) | Intuitive Machines, Skycorp |
| Microsoft Azure Space | Infrastructure | Software & Ground Station (Azure Orbital, Space Edge) | Thales Alenia Space, SpaceX, SES |
| OrbitsEdge | Startup | Ruggedized Hardware (SatFrame protection for COTS servers) | HPE, Vaya Space |
| Ramon.Space | Hardware | Radiation hardened electronics (DSP/Storage processors) | LEOcloud, Space Agencies |
| ReOrbit | Infrastructure | Software-Defined Satellites (Networking & Data Flow) | ESA, Thales Alenia Space |
| Sidus Space | Infrastructure | Edge AI Satellite Bus (LizzieSat / FeatherEdge) | NASA, Xiomas |
| Space Compass | Network | Optical Data Relay (Space Integrated Computing Network) | NTT, SKY Perfect JSAT |
| SpaceX | Infrastructure | Orbital Compute Capability (Starshield/Starlink edge processing) | US Gov |
| Starcloud | Startup | AI Training/Inference (Launched H100 payload; fmr. Lumen Orbit) | Crusoe, Nvidia |
| Thales Alenia Space | Infrastructure | Payload Integration (Space Edge Computing) | Microsoft |
| TransAstra | Logistics | Orbital Logistics (Space tugs for datacenters) | Space Force, ThinkOrbital |
| Ubotica | Hardware | AI Hardware Acceleration (CogniSat / LIVE Earth Intelligence) | ESA, Open Cosmos |
| Unibap | Hardware | Edge Computing Platform (SpaceCloud / iX5 & iX10) | ESA, D-Orbit |
| Voyager | M&A Rollup | IaaS / Cloud Services (RedHat OpenShift in orbit) | Axiom Space, Microsoft, Ramon.Space |
While I believe this list will grow, it is unclear how fast that growth will be. Still, the domain is exciting to follow and I look forward to providing updates. Eventually, I’ll organize all the Key Partners in this list and track the M&A along the way.
Enjoyed this post?
Consider supporting my sponsor!
🔓 Unlock Your Best Self with Lida Coaching! 50% off regular coaching rates until July 31, 2026!
A quick message from Lida:
With 18 years building, marketing, and scaling products at startups, I work with founders and leaders stuck between where they are and where they need to be. I help you cut through the noise, make sharper decisions, and get your team executing with clarity instead of chaos.
Most founders mistake complexity for strategy. They keep adding when they should be subtracting. When I'm brought in, the team is usually doing a lot right, and the real issue is focus and direction. I diagnose what's actually going wrong across product, positioning, and execution, then reset direction and get everyone aligned around one clear path forward that will lead to growth.
A bit of what I've done: I was the first PM on a fintech product that hit 10,000+ users within months, built a national incubator that supported over $500M in startup funding, and led teams at 23 Design before it was acquired by frog (the design team behind Apple’s first products). I've spoken at Google, Microsoft, and LinkedIn, etc.
🔓 Unlock Your Best Self with Lida Coaching! 50% off regular coaching rates until July 31, 2026!
