Orbital Data Centers

Launch Cost Trajectory

Historical and projected $/kg to LEO: from Space Shuttle ($54,500/kg) through Falcon 9 ($2,600/kg) to Starship targets ($100-200/kg). Google's $200/kg parity threshold marked.
Active/OperationalProjected/TestingIn DevelopmentRetired
Updated at 2026-03-18

Space vs Terrestrial Cost Comparison

1 GW orbital DC costs ~$42.4B vs $14.8B terrestrial (McCalip analysis). Space wins on energy, cooling, water; loses on launch cost, maintenance, and insurance.

Orbital (Space)

  • Total Cost (1 GW)
    $42.4B

    McCalip Calculator, TechCrunch

  • LCOE (Levelized Cost of Energy)
    $891/MWh

    McCalip Model

  • PUE (Power Usage Effectiveness)
    1.0 (theoretical ideal)Advantage

    TechTarget, AirSys

  • Cooling Cost (% of OpEx)
    0% (radiative cooling to space)Advantage

    DataSpan, Thunder Said Energy

  • Water Consumption
    ZeroAdvantage

    Lawrence Berkeley Lab

  • Solar Irradiance
    1,361 W/m²Advantage

    NASA GSFC, Wikipedia

  • Solar Capacity Factor
    ~95% (LEO sun-sync)Advantage

    McCalip, SatNews

  • Google Solar Advantage
    8x annual energy productionAdvantage

    Google Research Blog

  • Land Use
    Zero land footprintAdvantage

    Industry consensus

  • Grid Connection
    Not needed (self-powered)Advantage

    Industry analysis

  • Maintenance Access
    Zero (no servicing capability)

    Gartner, IEEE Spectrum

  • Hardware Replacement
    Impossible (deorbit and replace)

    Gartner

  • Latency to End Users
    20-40 ms (LEO round-trip)

    Industry standard

  • Scalability
    Limited by launch cadence

    Deutsche Bank

  • Insurance/Risk
    No mature space DC insurance market

    DCD

  • Environmental Impact
    Zero emissions (solar), debris riskAdvantage

    ESA, IEA

Terrestrial (Ground)

  • Total Cost (1 GW)
    $14.8BAdvantage

    McCalip Calculator, TechCrunch

  • LCOE (Levelized Cost of Energy)
    $398/MWhAdvantage

    McCalip Model

  • PUE (Power Usage Effectiveness)
    1.54 (US average)

    TechTarget, AirSys

  • Cooling Cost (% of OpEx)
    30-40% of energy consumption

    DataSpan, Thunder Said Energy

  • Water Consumption
    17B gallons/year (US DCs, 2023)

    Lawrence Berkeley Lab

  • Solar Irradiance
    ~200-300 W/m² effective

    NASA GSFC, Wikipedia

  • Solar Capacity Factor
    23.5% (US average)

    McCalip, SatNews

  • Google Solar Advantage
    Baseline

    Google Research Blog

  • Land Use
    Hundreds of acres per GW-scale DC

    Industry consensus

  • Grid Connection
    Major bottleneck (years to connect)

    Industry analysis

  • Maintenance Access
    Full 24/7 accessAdvantage

    Gartner, IEEE Spectrum

  • Hardware Replacement
    Standard IT refresh cyclesAdvantage

    Gartner

  • Latency to End Users
    <1 ms (on-premises fiber)Advantage

    Industry standard

  • Scalability
    Limited by power/land/permits

    Deutsche Bank

  • Insurance/Risk
    Standard commercial insuranceAdvantage

    DCD

  • Environmental Impact
    CO2 emissions, water use, land use

    ESA, IEA

Updated at 2026-03-18

Terrestrial Data Center Market Context

$61B invested in DC construction (2025), 1,189 hyperscale DCs globally, 415 TWh energy consumption (2024), ~$443B hyperscaler capex (2025).
MetricValueYearTrend / ContextSource
Global DC Construction Investment$61B2025Record year; doubling from $30B in 2023CNBC, S&P Global
DC Construction Market Size$241B2024Growing at 11.8% CAGR to $457B by 2030Grand View Research
DC CapEx Pipeline (projected doubling)$1.1 trillionBy 2029Doubling from ~$430B in 2024 baseeWeek
Hyperscale DCs Worldwide1,189Q1 2025Growing steadily; US accounts for 54% of capacitySynergy Research
Global DC Energy Consumption415 TWh20241.5% of global electricity; 15% annual growthIEA
Projected DC Energy Consumption945 TWh2030Doubling in 6 years driven by AIIEA
US DC Power Demand Growth+22% in 20252025Tripling to ~150 GW by 2028-2030S&P Global
US DC Share of Electricity4.4% today → 12% by 20302025-2030From ~38 GW to 134 GWWorld Resources Institute
Grid Connection Queue10,300 projects / 1,400 GW capacityEnd 2024Avg 3-7 year wait; $10-50M+ substation upgradesEngineering News-Record, LandGate
Delayed DC Projects36 projects / $162B investment blocked/delayedMid-2025Grid bottleneck is primary constraintData Center Frontier
Top 5 Hyperscaler CapEx~$443B (Amazon $125B, Microsoft $118B, Google $93B, Meta $72B)2025Projected to grow to ~$602B in 2026MUFG, Bloomberg, Axios
US DC Water Consumption17B gallons/year (449M gal/day)2023Single 5M gal/day facility = 10% of county water supplyLawrence Berkeley Lab, EESI
Average PUE (industry)1.562024Down from 2.5+ in 2007; plateauing; Google at 1.09Statista, Google
Land ConstraintsSilicon Valley near $100/sq ft2025⅔ of new capacity moving outside NoVA and SVJLL
Updated at 2026-05-06

Power Advantage: Space Solar

Space solar: 1,361 W/m², 95%+ capacity factor, 8x more productive than Earth. Solar array specific power evolution from ISS (27 W/kg) to advanced thin-film (>200 W/kg).

Space (Orbit)

  • Solar Irradiance (constant)1,361 W/m² (solar constant)

    Wikipedia, PVEducation

  • Atmospheric Absorption Loss0% (no atmosphere)

    Wikipedia Solar Irradiance

  • Effective Average Irradiance1,293-1,361 W/m² (LEO sun-sync)

    NASA GSFC, S&P Global

  • Capacity Factor~95%+ (sun-synchronous orbit)

    McCalip Analysis, SatNews

  • Night/Weather Downtime0-5% eclipse (orbit-dependent)

    Orbital mechanics

  • Annual Energy Output (per m²)~11,200 kWh/m²/year

    Calculated from irradiance × capacity factor

  • Google's 8x Productivity Claim8x more power per panel per year vs Earth

    Google Research Blog

  • ISS Original Solar Arrays27 W/kg specific power

    NASA Solar Power Technologies

  • ISS iROSA Arrays (Redwire)75.3 W/kg specific power

    Wikipedia ROSA, Redwire

  • Advanced Thin-Film (target)150-250 W/kg specific power

    ScienceDirect

  • ISS Total Solar Power~120 kW (end-of-life with iROSA)

    NASA ISS Facts

  • Starcloud Solar PowerNot publicly disclosed

    Starcloud

  • Cost of Space Solar Panels~$500-1,000/W (current)

    IEEE Spectrum, industry estimates

  • Degradation Rate~1-2% per year (radiation)

    Industry standard

Earth (Surface)

  • Solar Irradiance (constant)~1,000 W/m² peak (sea level, clear day)

    Wikipedia, PVEducation

  • Atmospheric Absorption Loss~25% absorbed/scattered

    Wikipedia Solar Irradiance

  • Effective Average Irradiance~200-300 W/m² (location-dependent average)

    NASA GSFC, S&P Global

  • Capacity Factor23.5% (US national average)

    McCalip Analysis, SatNews

  • Night/Weather Downtime50-75% (night + weather + seasons)

    Orbital mechanics

  • Annual Energy Output (per m²)~1,400 kWh/m²/year (good locations)

    Calculated from irradiance × capacity factor

  • Google's 8x Productivity ClaimBaseline 1x

    Google Research Blog

  • ISS Original Solar Arrays-

    NASA Solar Power Technologies

  • ISS iROSA Arrays (Redwire)-

    Wikipedia ROSA, Redwire

  • Advanced Thin-Film (target)-

    ScienceDirect

  • ISS Total Solar Power-

    NASA ISS Facts

  • Starcloud Solar Power-

    Starcloud

  • Cost of Space Solar Panels~$0.20-0.50/W (terrestrial)

    IEEE Spectrum, industry estimates

  • Degradation Rate~0.5% per year (weather/UV)

    Industry standard

Updated at 2026-03-18

The Cooling Challenge

ISS rejects 70 kW via 422 m² radiators. A 1 GW orbital DC at 40% efficiency would need ~834,000 m² of radiator area. Liquid droplet radiators offer 7x improvement.

ISS Active Thermal Control (EATCS)

70 kW heat rejection

External Active Thermal Control System using ammonia loops + radiators

Wikipedia EATCS, NASA

ISS Total Radiator Area

422 m²

14 radiator panels on station truss

Wikipedia EATCS

ISS Radiator Power Density

~166 W/m²

70 kW / 422 m² = 166 W/m²

Calculated

Space Ambient Temperature

~2.7 K (-270.5°C)

Cosmic microwave background; near absolute zero

Physics standard

Heat Rejection Method in Space

Radiation only (Stefan-Boltzmann law)

No convection or conduction possible in vacuum

Thermodynamics

1 MW GPU Cluster Waste Heat

~600-700 kW (at 60-70% efficiency)

Must be radiated; cannot use air or liquid to outside

Industry estimate

Radiator Area for 1 MW

~2,500 m²

At ~400 W/m² (optimistic advanced radiators)

Space Computer Blog

1 GW DC Waste Heat (40% efficiency)

600 MW

60% of total power becomes waste heat

Medium Analysis

Radiator Area for 1 GW (600 MW)

~834,000 m²

At ~166 W/m² ISS-class radiators (834,000 m² ≈ 83 hectares)

Medium Analysis

Radiator Mass for 1 GW

~2,250 tonnes

At typical ~2.7 kg/m² radiator mass

Medium Analysis

Launch Cost for 1 GW Radiators

~$450M (at $200/kg)

Just for radiator mass to orbit

Calculated

Advanced Liquid Droplet Radiators

10x lighter than solid radiators

Research-stage technology; sprays droplets to radiate heat

Wikipedia LDR, ScienceDirect

Starcloud Approach

Distributed micro-satellites

Each small satellite has modest thermal load; avoids mega-radiator problem

Starcloud strategy

Google Suncatcher Approach

81-satellite cluster

Distributes computing and thermal load across many small spacecraft

Google Research

Updated at 2026-03-18

Launch Provider Comparison

10 vehicles compared: Falcon 9, Falcon Heavy, Starship, Electron, Neutron, New Glenn, Vulcan, Ariane 6, Long March 5, and Long March 9.
VehicleOperatorPayload to LEO (kg)Total CostCost per kgReusabilityStatus
Starship (50-70 flights)SpaceX150,000~$2-3M$13-20Full reuse targetProjected (2030s)
Starship (20 flights)SpaceX150,000~$5M$32.50High reuseProjected
Starship (6 flights)SpaceX150,000~$12-14M$78-94Partial reuseProjected
Starship (single-use)SpaceX150,000-200,000~$90M (est.)$250-600Expendable configTesting (2025)
Falcon 9 (internal/marginal)SpaceX22,800~$14.3M (internal)$629Booster reuse (high-flight)Active (Starlink deploys)
Falcon HeavySpaceX63,800~$97M$1,400Side booster reuseActive
Long March 9 (projected)CASC (China)150,000~$225M (est.)~$1,500Partially reusable (planned)First flight 2033
New GlennBlue Origin45,000~$68M$1,511Reusable first stageFirst launch Jan 2025
Falcon 9 (customer)SpaceX22,800~$67M (list price)$2,600Booster reuse (20+ flights)Active (workhorse)
Long March 5CASC (China)25,000~$75M~$3,000ExpendableActive
Neutron (projected)Rocket Lab13,000$50-55M~$4,000Reusable first stageIn development (2026)
Ariane 6 (A64)ArianeGroup21,600~$115M (target)~$5,324ExpendableActive
Ariane 6 (A62)ArianeGroup10,350~$80M (target)~$7,729ExpendableActive (first launch Jul 2024)
Ariane 5ArianeGroup21,000~$178M$8,476ExpendableRetired 2023
Vulcan CentaurULA10,800$110M$10,185Expendable (SMART reuse planned)Active (first launch Jan 2024)
Atlas V (401)ULA10,986~$130M$11,837ExpendableActive (retiring)
Delta IV HeavyULA28,790~$350M$12,157ExpendableRetired 2024
ElectronRocket Lab200-300$7.5M$25,000Expendable (Neutron: reusable)Active
Space ShuttleNASA27,500~$1.5B per mission$54,500Partial (orbiter + SRBs)Retired 2011

Orbital DC Viability Threshold: $200/kg

Target for orbital DC viability. Vehicles below this cost enable economically viable space data centers.

Updated at 2026-03-18

Ask Sterling

Register for a premium account to gain access to Sterling AI.

Get Started

Things you can ask Sterling:

Summarize Tesla's latest earnings reportWhy did NVIDIA's margins expand?Compare Apple vs Microsoft's cash flowWhat's driving the EV sector growth?
Orbital Data Centers - Economics | Sterling