Chinese AI Data Centers Still Specify Nvidia Chips Under US Curbs

Reuters reported on November 13, 2026 that its review of more than 150 Chinese procurement notices, investment plans, construction contracts and corporate disclosures identified over 23 billion yuan ($3.2 billion) in computing projects and equipment agreements designed around Nvidia processors restricted by US export controls.

The total spans planned investments, signed contracts and completed installations. It covers buildings, power systems and other infrastructure, making it a measure of the broader buildout rather than spending on Nvidia chips alone. Projects stretch from coastal technology hubs to remote western provinces and extend beyond major technology companies to local government-backed ventures, universities and regional computing centers.

Nvidia has said US controls effectively shut it out of China’s AI data-center market and has excluded China data-center computing revenue from its forecasts, citing regulatory uncertainty. The documents show Chinese customers continuing to plan purchases and, in some cases, operating restricted processors, so exclusion from direct sales has not removed Nvidia hardware from Chinese project designs.

Planned and Operating Facilities

Among the proposed facilities in the companion project factbox, a 6.2 billion yuan AI computing center in Xianyou county, Fujian, seeks commercial partners and specifies H100 chips for large-scale AI development. In Shaanxi, a Sinomach subsidiary won a contract tied to a 3 billion yuan computing-center project specifying H100 and H200 processors.

A 2.26 billion yuan project in Jilin combines H100 processors with Chinese-developed chips, and its developer expects bank loans to finance about 80% of the investment. An Inner Mongolia development valued at 1.5 billion yuan proposes 625 H200 servers.

A local company awarded a 1.16 billion yuan construction contract in Gansu for a computing center that plans approximately 3,072 H100 and H200 processors. The tender documents cited in the review do not establish whether the processors were delivered, leaving a gap between the facility’s design and its confirmed hardware.

Other facilities are already using restricted hardware. Wuxi’s municipal government said in March 2025 that an H800 cluster had entered operation in Jiangsu. Damingcheng disclosed that a Fujian venture had built an H800 cluster and signed a computing-services agreement originally worth 759 million yuan, although the contract ended early.

Export Controls Expose Procurement and End Use

Washington began restricting Nvidia’s advanced AI processors in 2022 and later tightened controls on modified chips developed for China, including the H800. The Trump administration has also moved to review H200 export licenses for eligible Chinese customers individually, subject to conditions. For facilities still seeking hardware, the review identifies potential delays, supply disruptions and costly redesigns around domestic chips if controls tighten further.

The documents do not establish how the restricted hardware reached Chinese users, and the projects have not been linked to Megaspeed or the allegations involving Jensen Huang, who resigned as Nvidia’s chief executive on November 12, 2026. Megaspeed has denied illegally diverting Nvidia chips to China, and there has been no public finding that Huang violated export-control laws.

Sunny Cheung, an analyst at the Washington-based Jamestown Foundation, said the findings reinforce his own view on enforcement: “Washington must stop focusing solely on who buys the chips on the surface and start examining who ultimately uses the computing power—and for what purpose.”

The project records separate two questions that statements about direct sales cannot resolve: whether announced facilities can secure the processors in their designs, and who uses the restricted computing capacity already installed.