Starcloud can’t put a data centre in space at $8.2M in one Starship

Overall sentiment

• Majority of commenters see orbiting data centers as technically possible but economically and operationally absurd for the foreseeable future.
• The idea is widely grouped with hype projects (Solar Roadways, Hyperloop, SpinLaunch), and some call it “VC fodder” or a future Theranos-style grift.
• A minority argues that if fully reusable launch really becomes cheap, some version of “compute in space” might eventually make sense, especially for in‑space workloads.

Power & cooling

• Proponents:
  • Near‑continuous solar in dawn–dusk sun‑synchronous orbits gives highly reliable, predictable power with minimal batteries.
  • Radiative cooling via large radiators is standard practice on satellites; ISS shows it works.
• Critics:
  • For data‑center‑scale loads (tens of MW to GW) radiator area and heat transport plumbing become enormous; heat pipes and coolant mass were not fully accounted for in the whitepaper or external analysis.
  • Space is a great insulator; without convection, dumping 40MW of heat is non‑trivial and likely heavier and more complex than on Earth.

Maintenance, reliability & robotics

• Terrestrial DCs see constant but manageable hardware failures; most are cheap to fix with human techs.
• In orbit, replacing parts needs either complex, redundant robotics or regular human servicing missions, both adding huge mass, cost, and complexity.
• Some suggest simply over‑provisioning and letting hardware “die in place” then replacing whole modules or entire satellites on multi‑year cycles.
• Radiation (SEUs, total dose) is a significant new failure mode, implying ECC everywhere and some shielding.

Economics & launch assumptions

• Napkin analyses hinge on optimistic Starship pricing ($250–1000/kg or lower). Some argue current article overstates launch costs; others note Starship is not yet operational at promised performance.
• Even if launch gets cheap, critics say almost every claimed advantage (power cost, isolation, latency‑tolerant training) can be achieved more cheaply with terrestrial solar + batteries or putting DCs in cold locations or underwater.

Legal, security & “outside jurisdiction”

• Multiple comments debunk the “no laws in space” fantasy: treaties make states responsible for all spacecraft they authorize, and operators remain subject to their home jurisdictions.
• Governments can target ground stations, people, or even destroy satellites if sufficiently motivated.

Use cases & future in‑space compute

• Some see potential long‑term only for:
  • On‑orbit processing of space‑generated data (imaging, sensors).
  • Relay / caching networks for deep‑space missions (e.g., Mars), though many argue these should be built when actually needed.
• Speculation exists around military use or shady “bulletproof hosting,” but commenters think political and technical realities make this niche and fragile.

Environmental & orbital risks

• Launch emissions and orbital debris are seen as significant downsides; large solar/radiator arrays increase collision cross‑section and ASAT vulnerability.
• “E‑waste reentry” (burning failed hardware in the atmosphere) raises pollution concerns.