Global satellite manufacturing output is forecast to reach $27.20Bn by 2030, expanding at a 15.3% CAGR over the 2024-2030 window as LEO constellation buildouts absorb the majority of new orders [S1].
The adjacent satellite communication market is on track to add USD 16,337.3M between 2025 and 2030 at a 10.5% CAGR, with year-on-year growth in 2026 printed at 9.5% and growth momentum characterised as accelerating, not decelerating [S2].
Constellation order book: SpaceX, Amazon Leo, and AST SpaceMobile drive near-term volume
Starlink set the cadence benchmark: nearly 70% of SpaceX's 115 Falcon 9 launches in 2024 were dedicated to constellation expansion, and a single November 2024 mission alone placed 23 Starlink satellites, 12 of which carried direct-to-smartphone payloads [S1]. Amazon is the closest challenger, with 212 Amazon Leo satellites on orbit by end of February 2026, a CapEx increase of $1B for 2026, 20 more launches planned inside 2026, and more than 30 earmarked for 2027, against a target of 700 Amazon Leo satellites by mid-2026 [S3].
AST SpaceMobile is converting proof-of-concept into a build phase, with company materials from late 2025 and early 2026 indicating planned launches every one to two months and a year-end 2026 target of 45-60 satellites in orbit [S4]. The BlueBird 6 design pairs a roughly 2,400 sq ft phased array with peak downlink targets up to 120 Mbps, sized for direct-to-standard-phone service using partners' licensed cellular spectrum [S4]. Telesat, by contrast, holds a $1B backlog two years ahead of Telesat Lightspeed deployment, and Blue Origin's TeraWave is positioning for terrestrial-grade connectivity from space [S3].
Segment split: smallsats, LEO, and communication payload dominate 2026-2030 mix
IndustryARC sizes the manufacturing market by mass class as small satellites (up to 500 kg), medium (500-1000 kg), and large (above 1000 kg), with the smallsat bucket absorbing the bulk of constellation demand [S1]. The same report breaks orbits into LEO, MEO, Sun-synchronous, and GEO, with LEO receiving the majority of new orders through the forecast window [S1]. On the application side, communication satellites lead, followed by reconnaissance and other uses [S1].
Technavio's 2026-2030 satellite communication forecast slices the demand pool by component (services, equipment, software) and by end-user (commercial, government and defense, enterprise, maritime, aviation) [S2]. End-user demand across both studies converges on the same set of buyers: commercial broadband, defense and government, maritime, aviation, and direct-to-device cellular [S2]. For procurement engineers, the practical implication is that pressure transmitter and flow meter specs written for ground-based service do not apply to on-orbit hardware; radiation-qualified variants, sourced under the same project envelope, are the binding requirement.
Direct-to-device: the fastest-growing subsegment through 2030
AST SpaceMobile's funding model illustrates the capital intensity: $1.204B in cash and equivalents at the end of Q3 2025 against $706.6M of total debt, underwriting a constellation sized to offer continuous, rather than intermittent, service in priority markets from 2027 onward [S4].
Quilty Space research quoted in industry coverage put it plainly: the consensus is that the market can support two or three LEO broadband constellations on commercial merit, but political and sovereign demand is likely to fund twice that number [S3]. Independent satcom advisor Carlos Placido's NCAT analysis added that mega-constellations "struggle to reach even 10 percent of the addressable consumer and enterprise market," implying aggregate supply still trails demand even as multiple operators scale [S3]. For spec-driven buyers, the practical procurement question is no longer whether LEO capacity exists, but which operator's spectrum, latency, and ground-segment footprint matches a given service area.
Orbital compute: a second demand wave forming 2026-2030
Space data centers emerged as a material incremental demand vector during 2026. Google published first-of-their-kind radiation test results for its Trillium (v6e) TPU in June 2026: under a 67 MeV proton beam simulating sun-synchronous LEO, the shielded accelerator survived a full 5-year-equivalent dose with no permanent failures, with HBM irregularities appearing only at roughly 3x the mission dose and no hard failures up to the tested maximum of 15 krad(Si) [S5]. NVIDIA followed at GTC in March 2026 with the Space-1 Vera Rubin module and radiation-approved IGX Thor, built on commercial TSMC nodes with ECC, lockstep, and triple modular redundancy, rather than exotic rad-hard silicon [S5].
Capital is following. Starcloud (formerly Lumen Orbit) closed a $170M Series A at a $1.1B valuation in March 2026, with roughly $200M total raised and an FCC filing for an 88,000-satellite, approximately 20 GW constellation; Aetherflux raised at a reported $2B valuation in late March 2026; and Orbital, a five-month-old startup, filed in June 2026 for up to 100,000 satellites at approximately 10 GW [S5]. Starcloud-1 placed an NVIDIA H100 in orbit in November 2025 and ran LLM training on it, with multi-GPU Starcloud-2 scheduled for October 2026 [S5]. Terrestrial demand growth makes the case: Goldman Sachs projects data-center power demand rising 165% by 2030 versus 2023 levels, with RAND's April 2026 analysis estimating only roughly 82 GW of net available new capacity by 2030 against demand that is forecast to double to 66 GW by 2027 [S5]. The cost-to-orbit gate is still open: Google modelling puts the sub-$200/kg threshold enabling per-kW competitiveness in the mid-2030s, and McKinsey cites $500/kg as the structural cost-inflection for Starcloud-class models [S5].
Geography, standards, and sourcing posture
Geography matters less than orbit. North America and Asia-Pacific, led by the US, China, India, Japan, and South Korea, account for the largest share of procurement budgets through 2030 in both IndustryARC and Technavio coverage, with Europe (UK, Germany, France, Italy, Spain, Netherlands) a close third and a Middle East cluster (UAE, Turkey, Saudi Arabia) emerging as a meaningful buyer [S1][S2].
Standards posture is converging. The FCC's Supplemental Coverage From Space framework reduces uncertainty for satellite-to-cell partnerships, though approvals remain process-driven and may vary by market [S4]. For international filings, the International Telecommunications Union (ITU) processing queue for non-geostationary orbits remains the binding choke point: early-2020s filings for hundreds of thousands of NGSO satellites have moved from "paper" status to active deployment pressure [S3]. Axiom deployed the first dedicated free-flyer orbital data center nodes in January 2026 with 2.5 Gbps optical links, setting a near-term reference point for on-orbit compute procurement [S5]. For ground-segment and launch buyers, the industrial valve, pressure sensor, and PLC supply chain on which both satellite factories and launch ranges depend remains a longer-lead procurement problem than the satellite buses themselves; buyers are well-advised to lock ground-support instrumentation contracts in parallel with flight hardware.
What does not work, and who should not anchor a 2026-2030 plan to satellite demand
Three failure modes are visible in the data. First, launch cadence, not satellite demand, is the binding constraint: Starship's record through 12 flights sat at 7 successes, roughly 58%, with the V3/Block 3 debut in May 2026 achieving payload deployment but losing the booster, and even a single dedicated data-center constellation of 50+ launches per 100 MW sits further out than the hardware [S5]. Second, interference and spectrum coordination remain a documented downside risk for direct-to-cell, with regulators still process-driven on a market-by-market basis [S4]. Third, capital intensity and dilution: ASTS's convertible debt and potential equity issuance extend runway but reduce per-share upside if cadence slips [S4]. Buyers and investors who treat satellite demand as a pure end-demand story, and ignore launch and spectrum constraints, will misread 2026-2030 numbers materially.
Trackable signals to watch through 2027: Amazon Leo's mid-2026 700-satellite milestone and FCC extension decision; AST SpaceMobile's year-end 2026 45-60 satellite count; Starcloud-2's October 2026 multi-GPU launch; and the Google-Planet Labs Suncatcher prototype mission targeted for early 2027. For related reading on adjacent 2026 procurement maps, see the drone market share by manufacturer supplier map and the OLED manufacturer vendor map for spec-driven buyers.