Overview
- An economic calculator by Varda Space Industries’ Andrew McCalip estimates a baseline $51.1 billion for a large orbital AI installation versus $15.9 billion on Earth, echoing a separate technical review by Andrew Côté.
- Google’s Project Suncatcher, developed with Planet Labs, targets prototype launches around 2027 to test TPU survivability, thermal behavior, fault tolerance, and optical networking in orbit.
- SpaceX is positioned as the economic lever through launch scale and Starlink connectivity, with Starship’s fully reusable design projected to lower costs if realized.
- Technical constraints remain severe, including Earth‑link latency, radiation‑induced errors and degradation, unserviceable hardware, and large radiator mass for heat rejection.
- A small set of niche applications is seen as viable near term, with in‑orbit processing of Earth observation data, defense and intelligence use, and disaster response cited as the most credible candidates.