Nvidia has said US export controls have effectively shut it out of China’s AI data centre market. A Reuters review of more than 150 Chinese procurement notices, investment plans, construction contracts and corporate disclosures tells a different story about demand: it identified projects and equipment agreements with a combined reported value exceeding 23 billion yuan ($3.2 billion), all built around the chipmaker’s restricted processors.
The findings, published on Friday, November 13, come a day after Nvidia co-founder Jensen Huang resigned as chief executive amid a scandal over his relationship with a Chinese businesswoman US authorities suspect of intelligence ties to Beijing. Huang, a vocal advocate of easing export restrictions, said last year that Nvidia’s market share in China had fallen from 95% to zero, and the company has excluded China data centre revenue from its forecasts, citing regulatory uncertainty.
“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,” Sunny Cheung, an analyst at the Washington-based Jamestown Foundation, told Reuters.
What the Review Found
The projects stretch from China’s wealthy coastal technology hubs to remote western provinces, and they reveal how deeply Nvidia technology remains embedded in China’s AI expansion, according to Reuters. The demand reaches well beyond China’s largest tech firms, into local government-backed ventures, universities and regional computing centres.
The projects are designed around restricted Nvidia H100, H200 and H800 processors. While some facilities are already operational or under construction, operators are exposed to potential delays, supply disruptions or costly redesigns around domestic chips as export controls tighten.
An IDC report cited by Reuters in October estimated that Nvidia still accounted for 34% of Chinese AI accelerator shipments in the first half of 2026, even though the company’s August filing said it was effectively excluded from the country’s data centre computing market.
The Projects, Province by Province
A proposed 6.2 billion yuan ($873 million) AI computing centre in Xianyou county, in coastal Fujian province, seeks commercial partners and specifies Nvidia H100 chips for a facility intended to support large-scale AI development, according to the Reuters factbox.
In northwestern Shaanxi province, state-owned engineering group Sinomach said a subsidiary won a contract as part of a 3 billion yuan computing centre specifying Nvidia H100 and H200 processors. In northeastern Jilin province, an April government prospectus describes another 2.26 billion yuan project combining Nvidia H100 processors with Chinese-developed chips, with the developer expecting bank loans to finance about 80% of the investment.
In Inner Mongolia, a 1.5 billion yuan development proposes installing 625 Nvidia H200 servers, and Chinese industry reporting in June said construction had begun on two buildings in its first phase. In northwestern Gansu province, a local company awarded a 1.16 billion yuan construction contract in July for a computing centre that plans to install approximately 3,072 Nvidia processors, according to tender documents specifying H100 and H200 hardware.
Other facilities are already using restricted Nvidia hardware. In eastern Jiangsu province, the Wuxi municipal government said in March 2025 that a computing cluster using Nvidia H800 processors had entered operation. In Fujian, Shanghai-listed Damingcheng disclosed in June that a computing venture had built an H800 cluster, under a computing-services agreement originally worth 759 million yuan, although the contract ended early.
Smaller purchases show the demand extending to research institutions. Shenzhen University awarded a contract for three H100 processors in September, while US-blacklisted Northwestern Polytechnical University bought two H200 chips the following month, Reuters reported separately.
Why It Matters
Huang’s departure could deepen these vulnerabilities by weakening Nvidia’s lobbying leverage as US lawmakers press for tighter controls, Reuters reported. Washington began restricting Nvidia’s advanced AI processors in 2022 and subsequently tightened controls on modified chips developed for China, including the H800. In January, the Trump administration moved to review H200 export licences for eligible Chinese customers individually, subject to conditions.
The scrutiny has intensified since Huang’s resignation on Thursday. Nvidia’s board has launched an independent investigation, saying it had no reason to believe company secrets or US national security were compromised, and there has been no public finding that Huang violated export-control laws. Megaspeed, the Singapore-based Nvidia cloud partner at the centre of the diversion allegations, has denied illegally diverting Nvidia chips to China.
The findings also land as Congress moves to tighten the rules further. OpenAI has backed bipartisan bills to curb China’s access to AI chips, as GPTOnline reported, and the Senate Banking Committee has summoned Huang and Commerce Secretary Howard Lutnick to testify in December, as GPTOnline reported.
What the Review Does Not Establish
Reuters is careful about the limits of its findings. The $3.2 billion estimate includes planned investments and signed contracts as well as completed installations, and it covers buildings, power systems and other infrastructure, so it does not reflect spending on Nvidia chips alone. The figures also do not establish how the restricted hardware identified in the review reached Chinese users, and the tender documents do not establish whether the processors in some projects have been delivered.
The Chinese projects identified by Reuters have not been linked to Megaspeed or the allegations involving Huang. Separately, forensic analysis has found evidence that Chinese AI models carry fingerprints of export-controlled Nvidia Blackwell GPUs, as GPTOnline reported, but that analysis likewise does not establish where the chips were located or how access was obtained.
