China Launches Ambitious AI Satellite Constellation to Redefine Space Computing

May 17, 2025 — China has taken a bold step toward dominating space-based computing with the launch of 12 AI-powered satellites on May 14, 2025, marking the first phase of its “Three-Body Computing Constellation.” Developed by Zhejiang Lab and commercial partner ADA Space, this initiative aims to create a massive orbital supercomputing network that could rival Earth’s most powerful data centers. With plans to deploy 2,800 satellites by 2030, China’s Star-Compute Program seeks to achieve a staggering 1,000 peta operations per second (POPS), potentially surpassing ground-based supercomputers like the U.S.’s El Capitan (1.72 POPS). This article explores China’s objectives, the technology behind the constellation, its implementation, implications for the global AI ecosystem, strengths and weaknesses, and its potential impact on U.S. AI strategy.

China’s Objectives: A Space-Based Computing Revolution

China’s primary goal is to establish a space-based computing infrastructure that processes data in real time, reducing reliance on Earth-based data centers. Traditionally, satellites collect data and transmit it to ground stations, where less than 10% of the data is processed due to bandwidth and latency constraints. The Three-Body Computing Constellation aims to overcome these bottlenecks by enabling in-orbit data processing, supporting applications like astronomical observations, environmental monitoring, and cross-orbit communication. By scaling to 1,000 POPS, China seeks to lead the global space computing industry, positioning itself at the forefront of a new era in orbital AI and data management. This aligns with China’s national strategy to achieve AI leadership by 2030 and bolster its “New Infrastructures” initiative.

The Technology: AI-Powered Satellites with Advanced Connectivity

Each of the 12 satellites boasts a computing capacity of 744 trillion operations per second, contributing to a combined 5 POPS and 30 terabytes of onboard storage. Equipped with an 8-billion-parameter AI model, these satellites can process raw data directly in orbit, eliminating the need for extensive ground transmission. High-speed laser inter-satellite links, operating at 100 gigabits per second, enable seamless data transfer across the constellation, creating a neural network-like system in space. One satellite also carries a cosmic X-ray polarimeter, developed by Guangxi University and the National Astronomical Observatories, to detect gamma-ray bursts and coordinate follow-up observations. The satellites, built by ADA Space (Chengdu Guoxing Aerospace Technology) and supported by HiStarlink’s laser communication terminals, represent a shift from traditional sensing and communication satellites to fully integrated AI and computing platforms.

Implementation: A Collaborative and Scalable Approach

The launch, executed via a Long March 2D rocket from Jiuquan Satellite Launch Center, was a collaborative effort led by Zhejiang Lab, with contributions from ADA Space for satellite platforms and HiStarlink for optical terminals. Zhejiang Lab plans to deploy over 50 satellites in 2025, with the long-term goal of 2,800 satellites by 2030. The constellation’s phased deployment ensures iterative testing of capabilities like cross-orbit laser communication and real-time data processing. Partnerships with ground-based computing centers, such as those operated by SoftStone and Kepu Cloud, complement the orbital network, creating a hybrid infrastructure for AI-driven applications. This collaborative model, backed by the Zhejiang provincial government and international partners, underscores China’s commitment to scaling space computing infrastructure rapidly.

Implications for the Global AI Ecosystem

The Three-Body Computing Constellation could reshape the global AI ecosystem by introducing space-based cloud computing as a viable alternative to terrestrial data centers. With global data centers consuming over 1,000 terawatt hours annually—equivalent to Japan’s total electricity use—orbital computing offers energy efficiency through solar power and heat dissipation in space. This could reduce the carbon footprint of AI operations, addressing environmental concerns associated with ground-based infrastructure. Economically, the constellation could enable faster data processing for industries like satellite communications, disaster management, and scientific research, fostering new AI-driven applications. However, it also raises concerns about data sovereignty and security, as space-based AI could process sensitive information beyond national jurisdictions. Militarily, the ability to process data in orbit could enhance China’s surveillance and defense capabilities, potentially shifting strategic balances.

Strengths and Weaknesses

Strengths:

• Real-Time Processing: In-orbit computing eliminates ground transmission delays, enabling rapid decision-making for time-sensitive applications.

• Scalability: The planned 2,800-satellite network offers a path to unprecedented computing power, potentially surpassing terrestrial supercomputers.

• Energy Efficiency: Solar-powered satellites with space-based cooling reduce the environmental impact compared to Earth-based data centers.

• Strategic Advantage: The constellation positions China as a leader in space computing, with potential economic, scientific, and military benefits.

Weaknesses:

• High Costs and Complexity: Deploying and maintaining 2,800 satellites requires significant investment and logistical expertise, with risks of delays or technical failures.

• Security Risks: Orbital AI systems could be vulnerable to cyberattacks or interception, raising concerns about data integrity and espionage.

• Regulatory Challenges: International space law and data governance frameworks may complicate operations, especially for cross-border applications.

• Technological Maturity: While promising, the technology is in its early stages, and scaling to 1,000 POPS may face unforeseen technical hurdles.

Impact on U.S. Government AI Strategy

The U.S. has tested edge computing in space but lags behind China’s deployment of a purpose-built AI constellation. The Three-Body Computing Constellation could challenge U.S. dominance in AI and space technology, prompting a reevaluation of long-term strategies. The U.S. government may need to accelerate investments in space-based computing to counter China’s lead, potentially through public-private partnerships like those with SpaceX or NASA. Enhanced focus on securing orbital data and developing countermeasures against potential military applications of China’s constellation will be critical. Additionally, the U.S. could prioritize international collaborations to establish global standards for space-based AI, ensuring data security and interoperability. Failure to respond could cede strategic advantages to China, particularly in economic and defense domains, as space becomes a new frontier for AI competition.

Conclusion

China’s Three-Body Computing Constellation marks a paradigm shift in AI and space technology, aiming to create the world’s first orbital supercomputer. By leveraging advanced AI, high-speed laser links, and a scalable satellite network, China is poised to redefine data processing in space. While the initiative offers significant benefits for the AI ecosystem, its challenges—cost, security, and regulatory hurdles—cannot be overlooked. For the U.S., this development signals an urgent need to bolster its space computing capabilities and align its AI strategy to maintain technological and strategic leadership in an increasingly competitive global landscape.

Sources: SpaceNews, NewsX, Business Today

More AI News Articles