China’s Military-Linked University Acquires Nvidia H200 AI Chips Despite US Export Restrictions

A Chinese university with close ties to the country’s military has purchased advanced Nvidia artificial intelligence chips despite being subject to long-standing US export restrictions, raising fresh concerns about Washington’s ability to prevent sensitive computing technology from reaching institutions linked to China’s defence sector.

Northwestern Polytechnical University, based in Xi’an, reportedly purchased two Nvidia H200 processors in October 2026 for 874,700 yuan, equivalent to approximately $123,000. Procurement records also indicate that the university obtained access to powerful Nvidia chips through domestic computing-service providers, highlighting how restricted technology may remain accessible through indirect commercial channels.

The findings add to growing scrutiny of the global semiconductor supply chain as the United States attempts to limit China’s access to advanced computing capabilities that could support military research and artificial intelligence development.

Why Northwestern Polytechnical University Is on the US Blacklist

Northwestern Polytechnical University is one of China’s prominent defence-focused academic institutions and belongs to the group commonly known as the “Seven Sons of National Defense.” Its research includes areas related to military aircraft, missiles, satellites and other advanced technologies.

The US Commerce Department added the university to its Entity List in 2001 over concerns about its alleged involvement in missile proliferation and military-related research. Restrictions were tightened in 2019, making exports of controlled American technology to the institution subject to a policy of denial in most circumstances.

The latest procurement records indicate that the two H200 processors were purchased for the university’s management school from a Chinese supplier, Shaanxi Bosuan Network Technology, rather than directly from Nvidia.

The purchase raises questions about how a blacklisted institution was able to obtain hardware that generally requires stringent export-control compliance.

Computing Services Provide Another Route to Restricted Chips

The university’s reported purchases were not limited to the two processors acquired outright. Records also show that it obtained access to computing capacity containing advanced Nvidia chips through contracts with Chinese technology companies.

In January 2026, its cybersecurity school reportedly awarded two seven-month contracts for access to servers equipped with eight H200 chips each. The agreements, worth approximately 441,700 yuan and 441,000 yuan, covered artificial intelligence model fine-tuning and reinforcement-learning training.

The university also reportedly rented computing capacity containing Nvidia A100 and A800 processors. These chips are subject to US export restrictions on sales to Chinese entities.

Such arrangements illustrate a significant enforcement challenge: restrictions on direct hardware exports may not prevent institutions from accessing computing power through locally operated servers and service providers.

However, the procurement records do not establish where every processor originated, where the servers were physically located or whether any particular transaction violated applicable laws.

Other Chinese Universities Also Obtained Nvidia Technology

The investigation identified similar purchases involving other Chinese academic institutions.

Shenzhen University reportedly purchased three Nvidia H100 processors in September 2026, while Wuhan University of Technology acquired a server containing an H100 chip and separately purchased four A100 processors.

Beihang University, another institution associated with China’s defence-focused university network, reportedly bought four Nvidia A800 chips in April.

These purchases suggest that the issue extends beyond a single university. Procurement records reviewed in the investigation documented 26 transactions involving Chinese universities and research institutions that bought restricted Nvidia chips or rented computing services using such processors.

The findings point to a broader market in which domestic suppliers and intermediaries may help Chinese organisations obtain access to advanced American computing technology.

US-China Technology Rivalry Intensifies

The developments come as Washington and Beijing compete for leadership in artificial intelligence and semiconductor technology.

The Nvidia H200 is a powerful processor designed for demanding AI workloads, including training and operating advanced models. US policymakers have raised concerns that access to such hardware could strengthen China’s capabilities in military modernisation, intelligence analysis and weapons development.

Although the Trump administration began permitting certain H200 exports to China on a case-by-case basis in January 2026, those arrangements did not remove the separate restrictions applicable to organisations on the US Entity List.

Nvidia and the US Commerce Department’s Bureau of Industry and Security did not respond to requests for comment cited in the investigation.

The purchases could intensify calls for stronger chip-tracking measures, tighter scrutiny of intermediaries and additional safeguards around cloud-computing services.

Ultimately, the case highlights the difficulty of enforcing export restrictions in a global technology market. Washington’s challenge is not only to control direct shipments but also to determine how advanced processors move through commercial networks and who ultimately gains access to their computing power.