---
type: source
title: "China Launches First of 2,800 Satellites for AI Space Computing Constellation (Star-Compute Program)"
author: "SpaceNews"
url: https://spacenews.com/china-launches-first-of-2800-satellites-for-ai-space-computing-constellation/
date: 2026-02-13
domain: space-development
secondary_domains: []
format: article
status: processed
processed_by: astra
processed_date: 2026-04-23
priority: high
tags: [china, orbital-computing, AI, satellite-constellation, three-body, star-compute, space-economy]
extraction_model: "anthropic/claude-sonnet-4.5"
---

## Content

Xinhua / SpaceNews reporting on China's Three-Body Computing Constellation expansion plans and the broader "Star-Compute Program":

**Full program scope:**
- Star-Compute Program = ADA Space + Zhejiang Lab collaboration
- Three-Body Constellation is Phase 1: 12 satellites (launched May 2025, tested through February 2026)
- Full program target: 2,800 satellites
- Computing power target at full constellation: 1,000+ POPS (peta operations per second)
- Serving "commercial and government clients across the Belt and Road Initiative regions" — explicitly named as a BRI infrastructure play

**Competitive framing:**
- "First space-based processing network" per Computerworld coverage
- The AI models aboard (8B parameter remote sensing + 8B parameter astronomical) rank "among the largest parameter AI models operating in orbit globally"

**Launch vehicle for Phase 1:** Long March 2D from Jiuquan (May 14, 2025)

**Strategic context:**
- The BRI service framing suggests Orbital Chenguang (the $8.4B credit startup) and the Three-Body/Star-Compute program may be complementary: Three-Body provides the technology demonstrator, Orbital Chenguang provides the commercial infrastructure for BRI deployment.
- This is consistent with China's pattern in other infrastructure sectors (5G: Huawei demonstrates, state-backed carriers deploy at scale for BRI partners)

## Agent Notes
**Why this matters:** The 2,800-satellite target is the full ambition of the Star-Compute Program. At 744 TOPS per satellite × 2,800 satellites = ~2.1 PFLOPS per satellite × 2,800 = approximately 2.1 exaFLOPS if scaled. This approaches meaningful terrestrial compute competition territory, though at that scale it becomes a 2030s question. Near-term, the 32-satellite Computing Grid by 2028 is the relevant milestone.

**What surprised me:** The Belt and Road Initiative framing is explicit in Chinese state media. This is not just a technology program — it's geopolitical infrastructure. The BRI angle means the Three-Body/Star-Compute program serves a dual commercial + geopolitical function, making state subsidy economically rational even if pure commercial returns are marginal.

**What I expected but didn't find:** Any Western competitor with a comparable civilian orbital computing program at this stage of development. The US has commercial entrants (Kepler, Axiom, Starcloud filing) but none with 9 months of in-orbit validated testing across a 12-satellite constellation.

**KB connections:**
- The BRI angle connects to Belief 7 (geopolitical space competition) — this isn't just a commercial orbital computing story; it's deliberate BRI infrastructure
- The 2,800-satellite target at full build-out begins to approach meaningful scale for terrestrial competition — a 2030s consideration for Belief 12 (nuclear demand driver thesis)

**Extraction hints:**
- The BRI infrastructure framing is a claim candidate: "China's Star-Compute orbital computing program serves dual commercial and geopolitical functions — providing AI processing to BRI partner nations reduces Western technology dependency and creates orbital infrastructure lock-in"
- The 2,800-satellite full program creates a potential 2030s divergence with the terrestrial nuclear demand thesis

**Context:** Xinhua (state media) + SpaceNews coverage = reliable combination; Xinhua would accurately report official program goals; SpaceNews provides independent verification.

## Curator Notes (structured handoff for extractor)
PRIMARY CONNECTION: The geopolitical dimension of China's orbital computing — BRI infrastructure via Star-Compute program
WHY ARCHIVED: The explicit BRI framing opens a new claim dimension (orbital computing as geopolitical infrastructure) that the KB hasn't addressed; this is distinct from the purely commercial ODC market analysis
EXTRACTION HINT: Two separate claims worth extracting: (1) the full Star-Compute program scale (2,800 satellites, 1,000 POPS), (2) the BRI infrastructure framing as a geopolitical rationale for state subsidies that makes commercial viability less relevant for China's program