Adapted by:
Javier Márquez
WriterI've been in media for over a decade, but I've been marveling at the possibilities that technology brings us much longer. I believe we live in a world where the digital revolution is changing everything and that Xataka is the best place to write about it.
Alba Mora
WriterAn established tech journalist, I entered the world of consumer tech by chance in 2018. In my writing and translating career, I've also covered a diverse range of topics, including entertainment, travel, science, and the economy.
On May 14, a Long March 2D rocket lifted off from the Jiuquan Satellite Launch Center in Northwestern China with an ambitious goal: launching the first 12 satellites of the so-called “Three-Body Computing Constellation” into orbit. This marks the official start of what China presents as the first distributed supercomputer network in space. If successful, it’ll become a huge development in the effort to advance AI beyond traditional data centers.
A leap in scale from data centers to space. The constellation is part of the “Star Compute” program, which is led by Chinese research institution Zhejiang Lab in collaboration with company ADA Space. The first mission, designated “021,” successfully placed 12 intelligent satellites into the same orbit. The satellites are named after Chinese cities such as Neijiang, Haikou, and Taizhou.
Each satellite is equipped with an advanced computing system and laser connectivity capable of reaching speeds up to 100 Gb/s. This creates an interconnected network in orbit. According to the Chinese government, these satellites possess a total capacity of 5 peta operations per second (POPS) and 30 TB of onboard storage.
What are POPS? According to HKEXnews, 1 POPS is equivalent to one quadrillion (10^15) operations per second. SpaceNews reports that China’s goal is to deploy a satellite constellation of 2,800 satellites, with a total capacity of 1,000 POPS.
A lab in orbit. Each satellite is equipped with a space computer developed by Zhejiang Lab, with a processing capacity of up to 744 tera operations per second per unit. Additionally, they carry an AI model with 8 billion parameters, specifically designed to operate in orbit. This setup enables real-time data processing, eliminating the need to transmit all information back to ground stations.
Use cases. The satellites are intended to provide specialized services, such as gamma-ray detection and the creation of 3D digital twins of entire regions. They’ll also be able to carry out remote observations with onboard processing and monitor natural phenomena.
One satellite even includes an X-ray polarimeter developed by the Chinese Academy of Sciences and Guangxi University. It’s designed to detect gamma-ray bursts and trigger coordinated observations with other missions within seconds.
Moreover, according to Ada Space, the collected data will also be used for civilian applications, including emergency response, immersive video games, and smart tourism.
An orbital cloud under sovereign control. The motivation behind this deployment is a strategic ambition. China wants to establish a space supercomputer infrastructure that offers global coverage, energy efficiency, and low latency. Unlike traditional data centers, satellites can harness solar power continuously and don’t require active cooling systems.
More power and more satellites. Ada Space has already confirmed it’s developing a second generation of satellites that will provide even greater power. The goal is straightforward. China wants to validate the on-orbit distributed computing architecture, scale it up, and create a space cloud. This cloud would be capable of operating large-scale AI models without depending on ground infrastructure.
Images | China Aerospace Science and Technology Corporation
View 0 comments