Reuters identifies over $3 billion in Chinese AI projects tied to restricted Nvidia chips
The review spans proposed facilities, construction contracts and operating clusters. Project commitments do not always establish that processors have arrived.
Loading page…
The review spans proposed facilities, construction contracts and operating clusters. Project commitments do not always establish that processors have arrived.
Listen to this story
Reuters’ November 13 review finds that Chinese AI-computing projects worth more than $3 billion specify Nvidia processors subject to U.S. export controls, but the total is not a measure of chips delivered or capacity already running. The projects range from facilities still seeking partners to operating clusters, so the figures show continued demand without establishing how much restricted hardware is in use. Tighter controls could delay builds, raise costs or force redesigns around Chinese alternatives.
A proposed Fujian center in Xianyou is priced at 6.2 billion yuan ($873 million), specifies H100 chips and is seeking commercial partners.
A Shaanxi project valued at 3 billion yuan specifies H100 and H200 processors; a subsidiary of state-owned Sinomach won a related contract.
Wuxi’s municipal government said an H800 cluster was operating in March 2025, while Reuters also reported an H800 cluster built by a Fujian venture.
Chinese computing projects worth more than $3 billion are built around Nvidia processors subject to U.S. export controls, a Reuters review published November 13 found. The projects stretch from proposed facilities seeking partners to clusters already reported as operating—a distinction that matters when measuring how much restricted computing hardware is actually in use.
The findings contrast with Nvidia’s position that U.S. export controls have effectively shut it out of China’s AI data-center market. Reuters examined government prospectuses, tender documents and company disclosures covering projects across coastal technology hubs and western provinces.
A proposed AI computing center in Xianyou county, Fujian, carries a 6.2 billion yuan price tag, reported as $873 million. It specifies Nvidia H100 chips and is seeking commercial partners.
In Shaanxi province, state-owned engineering group Sinomach said a subsidiary won a contract connected to a 3 billion yuan computing-center project specifying H100 and H200 processors.
An April government prospectus describes a 2.26 billion yuan project in Jilin combining H100 processors with Chinese-developed chips. Its developer expects bank loans to finance about 80% of the investment.
In Inner Mongolia, a 1.5 billion yuan development proposes installing 625 Nvidia H200 servers. Chinese industry reporting in June said construction had started on two buildings in its first phase.
In Gansu, a local company awarded a 1.16 billion yuan construction contract in July for a center planning approximately 3,072 Nvidia H100 and H200 processors. A consortium led by a Shaanxi Construction Engineering subsidiary won the contract. The tender documents do not establish whether the processors have been delivered.
Other facilities have advanced further. The Wuxi municipal government said in March 2025 that a computing cluster using Nvidia H800 processors had entered operation.
Shanghai-listed Damingcheng disclosed in June that a computing venture in Fujian had built an H800 cluster. A computing-services agreement it signed, originally worth 759 million yuan, ended early.
For developers still building around restricted U.S. processors, Reuters identifies a supply risk: tighter controls could delay construction, inflate costs or force redesigns around domestic alternatives.
Loading discussion...
Join the conversation
Explain whether redesigning around domestic alternatives would be the better choice.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.