Which parts cap how fast the space economy can grow, who makes them, and what should decision-makers take from this study?
The short answer. satellites themselves are no longer the scarce item. SpaceX builds about 70 Starlink satellites a week and Amazon says it builds about 30 a week; in 2025 the world launched 4,492 payloads, ten times the 2019 figure. What is scarce is a short list of parts that only a handful of qualified suppliers make, that take a year or more to deliver, and that cannot be swapped without requalification. Setcoin's bottleneck scorecard ranks five at the top: optical inter-satellite terminals bought on the open market, solid rocket motors, launch slots, radiation-hardened processors and III-V multi-junction solar cells. Beneath the solar cells and radio parts sit two minerals, gallium and germanium, of which China produces 99% and most respectively; its ban on exports to the United States is suspended only until 27 November 2026. The largest constellation builders have escaped many of these limits by making parts themselves, which moves the bottleneck onto everyone else: defence programmes, European sovereign systems and smaller operators.
Ten findings
| # | Finding | Chapter |
|---|---|---|
| 1 | Satellite output has grown tenfold since 2019, almost all of it inside two vertically integrated companies | Chapter 3, Chapter 8 |
| 2 | The binding constraints are specific parts with few qualified suppliers, not satellites or factories in general | Chapter 44 |
| 3 | Merchant optical terminals rank first: the US defence constellation needed more than 500 and had 20 by January 2025 | Chapter 14 |
| 4 | Solid rocket motors are shared with missiles, and missile demand is winning the capacity | Chapter 19 |
| 5 | III-V solar cells are made in megawatts a year; mega-constellations need tens of megawatts, so they use silicon | Chapter 11, Chapter 12 |
| 6 | Germanium prices roughly doubled in 2025; the cell cost effect is real but small, the availability effect is large | Chapter 13, Chapter 48 |
| 7 | Small parts, switchgear and connectors, now have the longest lead times, up to 27 months | Chapter 22 |
| 8 | Workforce and test capacity are slower to fix than factories | Chapter 28, Chapter 31 |
| 9 | Vertical integration and acquisitions (Rocket Lab bought Mynaric and agreed to buy Iridium) are reshaping the merchant base | Chapter 51 |
| 10 | Closing the gaps to 2030 needs roughly $2–5 billion of targeted capacity, not tens of billions | Chapter 49 |
Key numbers
| Measure | Value | Source |
|---|---|---|
| Payloads launched worldwide, 2025 | 4,492 (453 in 2019) | Setcoin SATCAT analysis, 18 Oct 2026 |
| Share of 2025 payloads that were Starlink | 71% | Setcoin SATCAT analysis |
| SpaceX satellite production rate | ~70 a week | SpaceX registration statement, 2026 |
| Optical terminals for SDA Tranche 1 Transport | >500 needed; 20 delivered by Jan 2025 | GAO, Feb 2025 |
| Germanium metal price | $1,991/kg (2024) → $4,100/kg (2025 average) | USGS, 2026 |
| China's share of primary gallium | 99% | USGS, 2026 |
| Longest reported lead time | Space-grade switchgear, up to 27 months | PwC/AIA study, March 2026 (as reported) |
| Credentialed assembly workers a year vs hires | ~13,000 vs ~244,000 | RAND, Aug 2026 (as reported) |
The bottleneck scorecard in one table
Source table: Score (0–100) (10 rows, shown as an exhibit above)
| Rank | Component | Score (0–100) | Tier |
|---|---|---|---|
| 1 | Optical inter-satellite terminals (merchant) | 83 | Critical |
| 2 | Solid rocket motors | 81 | Critical |
| 3 | Launch slots | 77 | Critical |
| 4 | Radiation-hardened processors and FPGAs | 75 | Critical |
| 5 | III-V multi-junction solar cells | 75 | Critical |
| 6 | Space-grade power switchgear | 71 | Serious |
| 7 | Infrared focal plane arrays | 70 | Serious |
| 8 | Connectors and harnesses | 68 | Serious |
| 9 | GaN and GaAs RF parts | 67 | Serious |
| 10 | Actuators and nozzles | 64 | Serious |
Setcoin scorecard. lead time, supplier concentration, demand growth, difficulty of substitution and geographic exposure, each scored 1–5 and weighted 25/25/20/15/15. Method in Chapter 4; full ranking in Chapter 44.
Exhibit 1.2
Five parts score as critical; optical terminals rank first
The ten highest bottleneck scores (Setcoin scorecard) · score, 0–100
The five interactive models
| Model | Question it answers |
|---|---|
| Bottleneck score | Which parts rank as chokepoints when you change the weights? (Chapter 44) |
| Constellation bill of materials | Does component supply cover the satellites planned for a year? (Chapter 46) |
| Lead-time schedule risk | When must each part be ordered, and which one sets the launch date? (Chapter 47) |
| Mineral shock | What does a germanium price spike do to solar cell cost? (Chapter 48) |
| Electric propulsion propellant | How much xenon, krypton or argon does a constellation need, and at what cost? (Chapter 17) |
Each model can be downloaded as a spreadsheet with live formulas.
How the study is organised
| Part | Chapters | Content |
|---|---|---|
| I. Foundations | Chapter 2 to Chapter 8 | The bill of materials, demand from the satellite catalogue, the scoring method, supply tiers, vertical integration |
| II. Components | Chapter 9 to Chapter 25 | One chapter per critical part: electronics, solar, minerals, optical terminals, propulsion and propellants, motors, attitude control, power, small parts, RF, structures, sensors |
| III. Capacity | Chapter 27 to Chapter 32 | Satellite factories, test facilities, launch, workforce, qualification |
| IV. Policy and geography | Chapter 33 to Chapter 40 | Export control, critical minerals, tariffs, single points of failure, the US, Europe, China, other nations |
| V. Demand cases | Chapter 41 to Chapter 43 | Defence, mega-constellations, direct-to-device |
| VI. Synthesis and tools | Chapter 44 to Chapter 59 | Scorecard, bill-of-materials balance, lead-time risk, mineral shock, capacity build, consolidation, vendor map, checklist, risks, signals, method |
Appendices. A, component reference table; B, glossary; C, Setcoin's satellite-catalogue dataset.
Reading paths
| Reader | Suggested path |
|---|---|
| Satellite programme manager | Chapters 2, 4, 14, 22, 28, 47, 54 |
| Component supplier | Chapters 6, 8, 44, 46, 49, 51, 53 |
| Government and defence buyer | Chapters 19, 33, 34, 37, 38, 41, 49 |
| Capital provider | Chapters 1, 8, 44, 49, 51, 56, 57 |
| Newcomer | Chapters 2, 3, 6, 27, 61 |
How this study relates to the series
This is the seventh study in Setcoin's space research programme. Study S5 (launch economics) explains the launch capacity that this study treats as one bottleneck among several; Study S4 (defence space and Golden Dome) sets out the defence demand that competes for the same parts; Study S6 (direct-to-device) covers the large-array satellites that push RF and power supply hardest. Facts are as of October 2026; Setcoin calculations and illustrations are labelled.