Reuters finds more than $3 billion in Chinese AI projects built on restricted Nvidia chips

Despite years of US export controls, Chinese AI computing projects worth over $3.2 billion still explicitly reference Nvidia's restricted H100, H200, and H800 processors. A Reuters review published Nov. 13 identified 23 billion yuan in procurement notices and contracts across China that name these banned chips, suggesting demand persists regardless of restrictions.

The examined records fall into three categories: planned investments, awarded contracts, and operational facilities. This geographical spread spans China's coastal tech hubs and remote western provinces, including Fujian, Shaanxi, Inner Mongolia, and Gansu, indicating the issue is widespread rather than isolated.

Major Projects Continue

The largest identified project is a 6.2 billion yuan ($873 million) AI center in Fujian that specifically requires H100 processors. In Shaanxi, a 3 billion yuan project linked to state-owned Sinomach involves both H100 and H200 chips. A facility in Gansu was awarded a 1.16 billion yuan contract in July for 3,072 H100 and H200 processors.

Controls vs. Demand

Washington began restricting AI chip exports to China in 2022, with rules tightened multiple times to close loopholes.

For Nvidia, the immediate concern is compliance exposure rather than lost sales, as the company has already excluded China data-center revenue from its forecasts. Congressional calls for stricter controls are intensifying, with this $3 billion-plus paper trail likely to spur legislative action.

The broader implication is that export controls can restrict supply without eliminating demand. Chinese buyers continue designing infrastructure around hardware they cannot legally obtain, potentially leading to project delays, redesigns, or reassessments if supply becomes unreliable.