China Sets 2030 Target to Quadruple AI Computing Capacity

On Monday, China's Ministry of Industry and Information Technology (MIIT) unveiled a five-year strategy aimed at achieving 9,800 eflops of intelligent computing capacity by 2030, representing a fourfold increase from the 2,185 eflops noted at the end of June. The initiative includes a total investment in information infrastructure of 3.8 trillion yuan (approximately US$532 billion) from 2026 to 2030 and plans to establish AI computing clusters equipped with at least 100,000 accelerator cards. Additionally, it stresses the importance of adapting infrastructure to utilize domestic chips. By the end of July, the capacity had increased to around 2,450 eflops, as reported by the National Data Administration. This expansion builds on the 'East Data, West Computing' initiative launched in 2022, which reallocates power-intensive tasks to western regions. China has set up eight national computing hubs, 10 data-center clusters, and three coordination regions, with intelligent computing capacity growing 177% year-on-year by June, featuring 52 facilities with over 10,000 accelerator cards each.