China’s Ministry of Industry and Information Technology said in its latest information and communication industry development plan that the country aims to reach 9,800 EFLOPS of intelligent computing power by 2030, setting out a multi-year expansion path for national AI infrastructure.

The plan calls for the orderly deployment of intelligent computing clusters with capacities of 10,000 cards and above. It also proposes building out inference computing power, indicating a broader focus than headline capacity targets alone.

According to data cited from the National Data Bureau, China’s intelligent computing scale stood at about 2,450 EFLOPS, measured in FP16, as of the end of July. That implies roughly a fourfold expansion would be needed over the next four years to meet the 2030 target outlined in the plan.

The ministry also disclosed that intelligent computing power was 2,185 EFLOPS at the end of June, up 177% from a year earlier. Separately, the plan uses a 2025 baseline of 1,590 EFLOPS and sets the 2030 objective at 9,800 EFLOPS, equivalent to about 6.2 times that baseline.

The document also includes a cumulative investment target of 3.8 trillion yuan for information infrastructure across the information and communication industry. That total covers areas including communication networks and is not entirely earmarked for AI computing power. The ministry said 52 intelligent computing facilities with capacities exceeding 10,000 cards have already been established nationwide.

Why It Matters

The announcement adds policy clarity to China’s push to scale AI infrastructure at a national level. It matters because it links headline compute targets with deployment priorities such as large clusters, inference capacity, and domestic chip compatibility, giving the market a clearer read on where industrial support may concentrate.

For the broader digital asset sector, the significance lies less in immediate trading impact and more in infrastructure direction. Large state-backed compute expansion can influence how investors and builders assess AI-related crypto themes, particularly projects tied to compute networks, data infrastructure, and hardware-dependent ecosystems.