“What changed on Earth badly enough that orbital data centers became economically plausible?”
18 distinct claims from 28 evidence rows across 12 independent hosts, spanning 2014–2035: US and European electricity data, Google Project Suncatcher's system tests and 2027 prototype plan, and counterevidence on launch prices, radiators, radiation qualification and current orbital costs. Duplicate variants of the same Google Research page are consolidated so each claim appears once.
Space stayed hard to cool. Vacuum prevents convection, so all waste heat must leave by radiation: one cited analysis puts the radiator for a single 10-kW server at roughly 200 m², and a one-megawatt orbital system needs about 2,500 m² of radiator on top of 5,640 m² of photovoltaics. “Economically plausible” therefore means worthy of prototypes and conditional mid-2030s models — not cheaper or proven today.
| Claim | Figure | Publisher | Source |
|---|
Method: 28 evidence rows gathered across 12 independent hosts, consolidated to 18 distinct claims (trailing-slash and query-string variants of the same Google Research page merged; no exact URL backs more than 5 rows). Figures measure electricity demand (TWh), launch price ($/kg), link throughput, radiation dose and radiator area as stated by each source; full sentences shortened for space. Compiled November 2025.