Beijing sets ambitious AI infrastructure expansion goals through 2030
China's Ministry of Industry and Information Technology has unveiled a five-year plan targeting 9,800 eflops of intelligent computing capacity by 2030, requiring 3.8 trillion yuan in information infrastructure investment. The plan also calls for deploying clusters with 100,000 accelerator cards and adapting infrastructure to domestic chips. Current capacity stood at 2,185 eflops at the end of June, meaning a more than fourfold increase is needed.
The five-year blueprint, spanning 2026 to 2030, outlines a 3.8 trillion yuan investment to reach 9,800 eflops. It mandates building clusters with 100,000 or more accelerator cards and emphasizes adapting systems to locally produced chips. As of late July, capacity had already climbed to roughly 2,450 eflops, with 52 large-scale facilities operational by June.
This expansion continues the "East Data, West Computing" initiative from 2022, which relocates energy-intensive workloads to western regions with cheaper land and power. The strategy now relies on eight national computing hubs, ten data-centre clusters, and three coordination zones linking computing with electricity supplies.
This massive infrastructure push could reshape regional economic dynamics, potentially benefiting western provinces through new data-centre jobs and energy demand. For technology firms, the emphasis on domestic chips may accelerate local supply chains but could also raise costs or limit performance. Consumers might see faster, more capable AI services, though the heavy investment could strain public resources or concentrate technological power in state-aligned entities. The scale of expansion may also intensify global competition in AI capabilities.