Chinese AI giant SenseTime is launching a large-scale "Galaxy" project in partnership with nearly twenty chip and infrastructure equipment manufacturers. This is not just another initiative—it is a strategic response to tightening export restrictions from the United States and Beijing's drive for technological independence in the field of artificial intelligence.
As part of the project, five computing clusters are planned, each equipped with 10,000 Chinese AI accelerators. For comparison, this is comparable to the capabilities of leading global supercomputing centers, but based entirely on domestic components. In addition to computing power, the project includes the creation of a specialized token processing hub and support for 200 AI startups.
Notably, SenseTime has not disclosed either the construction timeline or the investment volume. This suggests that the project may be long-term and capital-intensive, and likely coordinated at the level of state strategy. Given that Chinese chips still lag behind Western counterparts from NVIDIA in performance, scaling up to 50,000 accelerators will require significant optimization of software and architecture.
Analytical commentary: This move by SenseTime sends a clear signal to the market: China is betting on sovereign AI infrastructures despite technological limitations. The large-scale deployment of clusters based on domestic chips could become a catalyst for the Chinese semiconductor ecosystem, but the key issues remain energy efficiency and the actual performance of these accelerators in training large language models. If the project proves successful, we will see a paradigm shift: dependence on Western GPUs will begin to weaken, and the Chinese market will receive a powerful boost for developing its own solutions.