Artificial intelligence competition now extends beyond Earth’s borders. China has taken the first step towards a new generation satellite network that carries computing capacity into space. The 12 satellites sent into orbit within the scope of the “Three-Body Computing Constellation” can not only collect data but also process it on their own. The long-term goal of the Beijing administration is to increase this structure to approximately a thousand satellites and establish a giant interconnected artificial intelligence and computing network in orbit. Thus, big data obtained from satellites can be analyzed in space without being brought down to Earth.
DATA WILL BE PROCESSED BEFORE REACHING THE EARTH
Most of today’s observation satellites first send the images they collect to stations on Earth. The new system developed by China aims to change this chain. Processors on the satellites will analyze the data while in orbit and send only the needed result back to Earth. Thus, huge amounts of raw data will not need to be constantly transferred to the earth. The first 12 satellites are not just observation instruments. Each of them has systems that can perform artificial intelligence calculations and data connections between satellites. The computational power targeted for the entire project is at 1000 peta transactions per second. China wants to build a distributed computing infrastructure in orbit large enough to compete with the world’s most powerful supercomputers.
ARTIFICIAL INTELLIGENCE IS NOW IN ORBIT
The Qwen-3 artificial intelligence model developed by Alibaba was also run on China’s orbital computing infrastructure. It took less than two minutes for the query sent from the ground to reach the satellite, be processed by artificial intelligence, and send the result back to Earth. Thus, it was shown that large artificial intelligence models can be operated directly in space. For China, the advantage is not just speed. The system will be able to generate a direct warning in orbit when fire, flood, ship movement or any unusual change is detected among the images coming from the satellites. This structure can significantly reduce the data transfer load in military observation, meteorology, maritime tracking and disaster monitoring systems.
GIANT NETWORKS
Star Hub’s uses include weather forecasting, disaster warnings, port and ship tracking, power grids and remote sensing. The Tiansuan network of 300 satellites is also being developed in China. Thus, several independent “space computer” projects are progressing simultaneously. One of the biggest problems facing giant artificial intelligence data centers around the world is the need for energy and cooling. It is aimed to benefit from solar energy in orbital systems for a longer period of time and to overcome some cooling problems with different methods. The decrease in launch costs also makes this idea, which previously did not seem economical, increasingly feasible.
NEW COMPETITION IN THE SKY
China’s step also points to a new phase of technology competition with the USA. However The biggest problem facing China is scale. Although it is possible for dozens of satellites to work together, increasing the system to hundreds or even thousands of satellites; Energy management is quite complex due to radiation, processor endurance, and orbital debris.
