Nvidia has told investors that US export controls have effectively shut it out of China's AI data-centre market. But a Reuters review of Chinese documents found more than $3 billion in computing projects across China still built around the company's restricted processors.
The review examined more than 150 procurement notices, investment plans, construction contracts and corporate disclosures, finding projects and equipment agreements worth more than 23 billion yuan, or about $3.2 billion. They stretch from wealthy coastal technology hubs to remote western provinces, reaching beyond China's largest tech firms into local government-backed ventures, universities and regional computing centres.
Billions in the pipeline
Across China, developers are advancing AI data-centre projects designed around Nvidia's restricted H100, H200 and H800 processors. In coastal Fujian, a proposed 6.2 billion yuan ($873 million) AI computing centre in Xianyou county specifies Nvidia H100 chips. In Shaanxi, state-owned Sinomach said a subsidiary won a contract on a 3 billion yuan centre specifying H100 and H200 processors. In Jilin, an April government prospectus describes a 2.26 billion yuan project combining H100 processors with Chinese-developed chips.
In Inner Mongolia, a 1.5 billion yuan development proposes 625 Nvidia H200 servers, with construction reported under way since June. In Gansu, a local company awarded a 1.16 billion yuan construction contract in July for a centre planning to install about 3,072 Nvidia processors, though it is unclear whether they ever arrived. Some are already running: Wuxi's municipal government said in March 2025 that an H800 cluster had entered operation, and Shanghai-listed Damingcheng disclosed in June that a venture had built an H800 cluster under a 759 million yuan services agreement that ended early. Shenzhen University bought three H100 processors in September, while US-blacklisted Northwestern Polytechnical University bought two H200 chips the following month.
A market Nvidia says it has lost
The findings land a day after Nvidia co-founder Jensen Huang resigned as chief executive amid a scandal over his relationship with a Chinese businesswoman suspected by US authorities of intelligence ties to Beijing. Huang, a vocal advocate of easing export restrictions, said last year Nvidia's market share in China had fallen from 95% to zero, and the company has excluded China data-centre revenue from its forecasts.
But the documents show Chinese customers still planning purchases and, in some cases, operating restricted processors. An IDC report cited by Reuters estimated Nvidia still held 34% of Chinese AI-accelerator shipments in the first half of 2026. Reuters noted the figures measure different things and do not show how the hardware reached users. The projects have not been linked to Megaspeed, the Singapore-based Nvidia cloud partner at the centre of the Huang allegations, or to Huang himself, and no public finding says Huang violated export-control laws.
Washington's dilemma
Washington began restricting Nvidia's advanced AI processors in 2022 and tightened controls on modified chips made for China, including the H800. In January, the Trump administration moved to review H200 export licences for eligible Chinese customers individually. The projects' operators now face delays, supply disruptions or costly redesigns around domestic chips — and Huang's departure could deepen those vulnerabilities by weakening Nvidia's lobbying leverage as US lawmakers press for tighter rules.
"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," said Sunny Cheung, an analyst at the Washington-based Jamestown Foundation.
Reuters cautioned that its estimate covers planned investments and signed contracts as well as completed installations — and buildings, power and other infrastructure, not chips alone. China's commerce ministry and Nvidia did not immediately respond to requests for comment.



