The United States eased regulations on Nvidia H200 chip exports to China, according to government and industry sources. The change came as part of an ongoing review of export controls on advanced semiconductor technology.
Under the revised rules, certain Nvidia H200 artificial intelligence chips may be shipped to Chinese companies that meet specific criteria. The policy adjustment expands the list of eligible buyers beyond those previously authorized. The revised process requires companies to apply for licences and meet screening requirements before receiving shipments.
Nvidia’s H200 chips are advanced graphics processing units designed for AI training and inference workloads. The chips support high performance computing tasks and are used in data centres and research environments. Export restrictions on advanced AI hardware formed part of U.S. efforts to limit China’s access to cutting edge technology.
The U.S. Commerce Department oversees export control policy and issues licences for technology shipments. The department said it will continue to enforce restrictions on certain high end chips and related components. However, some H200 variants may qualify for export under the revised criteria.
Under the new rules, eligible Chinese buyers must demonstrate compliance with U.S. national security and foreign policy considerations. Licence applications undergo review before approval. Companies that receive licences must abide by reporting and usage conditions as stipulated by regulators.
The adjustment affects Chinese cloud service operators and research institutions that rely on imported AI hardware. China’s tech industry previously faced limitations on acquiring top tier processors due to export controls introduced in recent years. The policy shift may ease procurement hurdles for some organisations.
Industry sources said Nvidia welcomed the regulatory change but did not disclose detailed corporate comment. Nvidia continues to work with regulators to navigate global trade and compliance obligations.
The export rule revision follows previous U.S. measures aimed at shaping the international semiconductor landscape. Export controls on advanced chips were introduced to protect sensitive technology and manage competition in AI and computing. The latest adjustments reflect ongoing evaluation of how technology flows affect economic and strategic interests.
Chinese companies seeking H200 chips must still apply for licences and undergo screening. Not all requested exports will be approved. The Commerce Department retains authority to deny licences based on security assessments.
The revised export controls apply specifically to Nvidia’s H200 series. Other high performance chips and related technology may continue to face stringent restrictions. The changes do not lift all curbs on China’s access to U.S. AI hardware.
Chinese tech firms and research groups are assessing how the eased regulations affect project planning and infrastructure development. The policy change may influence cloud services, AI research, and data centre investments within China.
The timeline for implementing the new rules was not immediately disclosed. Companies involved in licence applications are awaiting guidance on processes and timelines from U.S. regulators.
