# Orbital lifecycle economics and workload fit

Canonical: https://semiaifoundry.com/learn/atlas/orbital-lifecycle-economics/
Revision: 2026-10-07
Rights: https://semiaifoundry.com/ai-atlas/licenses.html

Frontier infrastructure · Atlas v1.3.0

# Orbital lifecycle economics and workload fit

Compares complete useful service under conditional launch and support assumptions.

[Explore in Atlas](https://semiaifoundry.com/ai-atlas/#orbital-lifecycle-economics)Payload electronics to fleet and lifecycle; conditional scenarios

Learning objectiveEvaluate conditional orbital economics without promoting launch assumptions into complete-system conclusions.

## The mechanism

An orbital computing proposal must pay for spacecraft, payloads, launch, ground infrastructure, operations, replacements and lost work. Solar energy availability can change operating inputs but does not erase those costs. Suncatcher’s launch-price analysis is explicitly conditional and incomplete as a total economic study; its projected future launch input must not be presented as a current quote. A fair terrestrial comparison uses the same accepted workload, quality and service targets. Space-generated data reduction and ground-user interactive services can produce very different communication costs. A useful scenario identifies the assumptions that allow its budget to close and any unresolved constraint that prevents a useful service.

1.  Choose a concrete workload and matching terrestrial service baseline.
2.  Close energy, thermal and link budgets before assigning sustained useful throughput.
3.  Include complete initial and recurring costs with a consistent lifetime and accounting boundary.
4.  Apply availability and accepted-result criteria without double-counting utilization already in measured throughput.
5.  Vary uncertain launch, mass, lifetime and service assumptions and identify where the preferred option changes.

## Science and assumptions

### Launch-only annualization

$$
C_{\mathrm{launch}}=\frac{p_{\mathrm{launch}}m}{L}
$$

Launch-only annualized cost per payload kilowatt is launch price per kilogram times mass per kilowatt divided by useful years.

Symbols and units

C\_launch

Launch-only cost annualized per payload kilowatt

currency/(kW·year)

p\_launch

Launch price per delivered kilogram under the selected scope

currency/kg

m

Launched mass per payload kilowatt

kg/kW

L

Useful operating life for the annualization

years

plaunch is currency/kg, m kg per payload kW and L useful years; result is currency/(payload kW·year).

Assumptions Only launch amortization; excludes spacecraft, ground, operations, financing, failures and other lifecycle costs.

## Engineering example

Synthetic $500/kg, 20 kg per payload kW and five-year life yields $2,000/(payload kW·year) for launch alone. Doubling mass doubles that component; no token rate or total ownership cost follows from it.

## Distinctions and tradeoffs

Comparing launch-only annualized cost to a terrestrial electricity bill is not a full system-cost comparison.

An attractive launch assumption can be outweighed by ground-data traffic, short lifetime, low availability or expensive thermal support.

AI workload fit determines useful results; no universal dollar-per-kW or MW-to-token factor can compare all terrestrial and orbital systems.

## Specifications and interfaces

Economic boundary

Complete lifecycle per accepted result

Launch-only cost is a component, not total cost of ownership.

Scenario status

Conditional assumptions and sensitivity

Future prices and fleet-scale availability require explicit uncertainty.

### Interfaces

Physical and operational budgets Input

Mass, power, thermal, links, lifetime and service quality

Conditional service economics Output

Cost per accepted result and unresolved bottlenecks

## Check your understanding

### Which denominator makes an orbital and terrestrial comparison meaningful?

1.  Only nameplate electrical power
2.  Only peak accelerator FLOPS
3.  Equivalent accepted workload results under matched conditions

Answer and reasoning

Equivalent accepted workload results under matched conditions

Useful service depends on quality, latency and availability, so identical power or peak arithmetic need not deliver identical outcomes.

## Evidence and scope

The example is synthetic. The source’s mid-2030s launch-price projection is forecast evidence, not a current market price or a completed feasibility demonstration.

No current calculator directly models this lesson’s main mechanism. Use the library for related bounds while retaining their stated omissions.

## Sources

-   [Towards a future space-based, highly scalable AI infrastructure system design, v2](https://arxiv.org/html/2511.19468v2)

[Published lesson data](https://semiaifoundry.com/ai-atlas/lesson-data/orbital-lifecycle-economics-5e1c75f1d1c56ba2.json) · [Atlas releases and licensing](https://semiaifoundry.com/ai-atlas/licenses.html)

## Continue exploring

### Foundations

-   [Launch, replacement and end-of-life mass](https://semiaifoundry.com/learn/atlas/orbital-replacement-mass/)
-   [Orbital optical links and contact budgets](https://semiaifoundry.com/learn/atlas/orbital-link-contacts/)

### Related lessons

-   [Orbital computing evidence ladder](https://semiaifoundry.com/learn/atlas/orbital-evidence-status/)