China launched 12 computing satellites in May 2025. These are early orbital computing networks, not a massive server farm floating above Earth yet. These satellites can process images and sensor data in orbit, then send the results back instead of downloading everything. China has also reported testing AI models, including Alibaba’s Qwen3, in space. Also, there is an Earth-based counterpart: the micro data center. Sometimes it’s just one server cabinet at a factory or remote site. The idea is similar: bring computing closer to where the data is created.[4][5] The U.S. isn’t sitting this out. American startup Starcloud launched an NVIDIA H100 GPU into orbit and demonstrated both running an AI model and a small training run.[3] So China has an early advantage in networked computing satellites, but that doesn’t mean it’s ahead in AI overall. These projects are still early, and neither country has a mature, massive orbital AI data center replacing the ones on Earth.[3][6] Some people wanted sources, so I included them here. [1] The data center space race has begun https://www.lightreading.com/satellite/the-data-center-space-race-has-begun [2] China Has 8-Billion-Parameter AI Running in Orbit as ... https://www.techtimes.com/articles/326180/20260901/china-has-8-billion-parameter-ai-running-orbit-shanghai-opens-space-computing-hub.htm [3] Data Centers in Space and the Future of Computing https://brightlio.com/data-centers-in-space/ [4] What Is a Micro Data Center? https://www.vertiv.com/en-us/about/news-and-events/articles/educational-articles/what-is-a-micro-data-center/ [5] What is an edge data center? https://www.cisco.com/site/us/en/learn/topics/computing/edge-data-center.html