Three of China’s most technology-intensive cities, Beijing, Tianjin, and Chengdu, have moved in rapid succession to establish dedicated institutions for space-based computing, signaling that China’s AI infrastructure ambitions are now extending well beyond terrestrial data centers and into orbit. The Global Times reports on the coordinated push, which unfolded over just a few weeks in late May and early June 2026, reflects a growing consensus among Chinese industry and government that space computing represents the next strategic frontier in the global AI race.
Space-based computing, the deployment of chips, servers, and data-processing hardware aboard satellites to process raw data in orbit rather than transmitting it back to Earth, is gaining urgency as ground-based infrastructure faces mounting constraints. Energy consumption, heat dissipation, and land availability are all becoming bottlenecks as AI demand surges. Orbital systems, powered by solar energy and capable of wide-area coverage, offer a potential path around those limits.
Beijing Establishes Its First Space Computing Innovation Center
Beijing moved first, establishing its first space computing innovation center in early June. The center will operate on a company-plus-alliance model, focusing on six key areas across the space computing industrial chain: chips, hardware, platforms, AI, networks, and applications. The goal is to build a fully coordinated industrial system that connects every layer of the stack.
Separately, GalaxySpace, a Beijing-based private space company, announced the establishment of a dedicated space computing research institute in Beijing E-Town. The institute will focus on space-borne computing chips, inter-satellite laser communications, space-based energy and thermal management, integrated space-ground networking, and space security standards, covering the core technical challenges that have so far limited the commercial viability of orbital computing.
Beijing’s space computing ambitions extend to hardware deployment: the E-Town institute has set a target of launching its first experimental satellite by 2028.
Tianjin Forms a National Consortium for Space Digital Intelligence
On May 29, at the World Intelligence Expo 2026 in Tianjin, the National Supercomputing Center in Tianjin joined with companies and research institutes to establish a joint consortium for space digital intelligence infrastructure. The consortium’s technical agenda spans the full hardware and software stack: from radiation-hardened on-board processors and scalable computing payloads to the software frameworks needed to manage distributed orbital workloads, the thermal and power systems that keep satellites operational, and the coordination protocols that link satellites to ground stations.
The Tianjin consortium’s formation at the World Intelligence Expo, China’s largest annual AI and intelligent technology showcase — underscores the degree to which space computing has moved from a niche aerospace topic to a mainstream AI infrastructure priority.
Chengdu Launches the World’s First Space-Based AI Cloud Platform
Chengdu has taken the most commercially aggressive step. On March 26, Chengdu-based aerospace company ADAspace, in collaboration with the MIIT’s small and medium-sized enterprises development promotion center, launched what it describes as the world’s first space-based computing power cloud service platform for enterprises, named Prometheus. Prometheus is designed to scale to 1,000 satellites by 2030, at which point ADAspace expects to begin commercial operations, with the vast majority of the constellation’s capacity dedicated to AI inference workloads rather than communications. A second phase would expand the network to 2,800 satellites by 2035.
If successful, Prometheus would represent the first fully AI-native orbital computing network at commercial scale, a significant milestone in the broader effort to move AI workloads beyond the physical constraints of ground-based infrastructure.
The Strategic and Commercial Case
The strategic rationale for China’s space computing push is clear. China’s terrestrial AI computing infrastructure has been expanding rapidly, with 42 intelligent computing clusters at the 10,000-card scale already built. But the energy and land demands of that infrastructure are growing faster than the grid can accommodate in many regions. Space-based computing offers a way to add capacity without the constraints of terrestrial deployment.
“The construction of space-based computing infrastructure is significant, as it will fundamentally alter the global supply pattern of computing power,” said Xiang Ligang, director-general of the Zhongguancun Modern Information Consumer Application Industry Technology Alliance. Xiang pointed to the physical ceiling that terrestrial data centers are approaching: the more AI demand grows, the harder it becomes to site, power, and cool new facilities on the ground. Orbital systems, drawing on continuous solar power and unencumbered by land constraints, offer a structurally different answer to that problem.
The China Academy of Information and Communications Technology (CAICT) estimates that China’s space computing industry will exceed 250 billion yuan ($36.6 billion) by 2030. Global market projections are even more ambitious, with industry analysts forecasting the worldwide space-based computing market will surpass $1 trillion by the same year.
Xie Lina, deputy director of CAICT’s Cloud Computing and Big Data Research Institute, said China is “already at the forefront globally in terms of engineering implementation and commercial space deployment” and ranks among the world’s top players in space computing exploration. She cautioned, however, that the industry remains in its early stages and requires a pragmatic approach and collaborative innovation to ensure steady growth.
The MIIT has pledged support for forward-looking research, calling for the gradual establishment of standards covering hardware, software, networking, and security, and promoting the development of space-borne radiation-resistant chips and inter-satellite laser communications.