Starship V3 Enables Orbital Data Centers

Starship V3 matters because it turns launch cost from a hard cap on space businesses into a variable that can keep falling with reuse and flight volume. Falcon 9 made satellite internet work. A fully operational Starship would make it cheap enough to loft entire racks of compute, power hardware, and cooling systems, which is why orbital data centers move from science project to possible product only at this cost tier.

• The economic step change is large. Falcon 9 has been modeled around roughly $2,700 per kg, while Starship V3 is framed around $10 to $100 per kg with 200 plus tons of payload. That is the difference between launching a satellite and launching the physical guts of a data center.
  1 sacra 2 sacra
• Launch price alone is not the whole story. SpaceX already has the orbital network piece through Starlink’s 9,000 plus satellites and laser links, and xAI gives it an internal anchor customer for training and inference demand. That vertical stack is what makes the data center thesis more than a rocket cost story.
  2 sacra 5 starcloud
• This is also about surplus capacity. SpaceX plans 100 plus Starship flights a year by 2028, and about 20 such flights would match the total mass put up by all 165 Falcon 9 launches in 2025. Once Starlink V3 deployment needs are covered, excess lift can seed entirely new markets.
  1 sacra 3 sacra 4 faa

If V3 proves out, the next phase is SpaceX using cheap lift to pull more of the space economy inside its own stack, first with bigger Starlink and Starshield payloads, then with compute infrastructure in orbit. That would shift SpaceX from selling launches and connectivity to owning the physical substrate for cloud, communications, and AI above Earth.