CHINA'S ORBITAL AI NETWORK GETS MUCH BIGGER

China is moving ahead with an ambitious plan to build a large space-based AI computing network, with partners around ADA Space and Zhejiang Lab aiming for a constellation that could eventually reach 2,800 satellites.

- 

The first 12 satellites were launched in May 2025 as the opening batch of the “Three-Body Computing Constellation,” and project materials say the network is intended to process more data in orbit rather than sending everything back to Earth first.

The core idea is straightforward: put more computing power in space so satellites can analyze imagery and other data where it is collected. China's state-linked reporting says this could reduce dependence on ground stations, cut transmission bottlenecks, and allow more real-time AI operations, especially in places where terrestrial connectivity is weak or unavailable.

THE 2,800-SATELLITE GOAL IS MUCH LARGER THAN THE CURRENT NETWORK

The long-term target is far bigger than what is already in orbit. Reports tied to the project describe the 12 launched satellites as the first pieces of a 2,800-satellite system linked to ADA Space's “Star Compute” program and the Three-Body Computing Constellation. The first batch is a proof-of-concept step, not the finished network.

China's official and state-backed reporting also describes a nearer-term expansion path. Xinhua says the constellation is planned to reach 100 satellites by 2027, while earlier Chinese reporting said the broader system would eventually scale much further. That means the current stage is still early, but the buildout is already moving from concept to hardware.

WHAT THESE SATELLITES ARE SUPPOSED TO DO