Nvidia is preparing to resume shipments of its H200 artificial intelligence chips to China as early as mid-February, according to sources familiar with the matter, signalling a calibrated adjustment to U.S. export controls while highlighting Beijing’s continued demand for high-performance computing hardware.
The planned deliveries would mark one of the most significant moves by Nvidia to re-enter the Chinese data-centre market following a series of tightening U.S. restrictions aimed at limiting China’s access to advanced AI and supercomputing technologies. The H200 chip, a successor to Nvidia’s widely used H100, is designed for training and deploying large language models and other compute-intensive AI workloads.
Sources said the version of the H200 destined for China has been adapted to comply with U.S. export rules, which impose caps on computing performance and interconnect speeds. Nvidia has previously redesigned products to meet regulatory thresholds, allowing it to continue supplying Chinese customers without breaching U.S. controls.
China remains a strategically important market for Nvidia, accounting for a meaningful share of its data-centre revenues prior to the introduction of export restrictions. Chinese cloud providers, research institutions, and technology firms have been racing to secure compliant AI accelerators as domestic demand for generative AI, industrial automation, and scientific computing continues to grow.
The expected resumption of shipments comes amid a complex geopolitical backdrop. Washington has sought to balance national security concerns with the commercial interests of U.S. technology companies, while Beijing has stepped up efforts to develop domestic semiconductor capabilities and reduce reliance on foreign suppliers. Despite these efforts, Chinese firms remain heavily dependent on Nvidia’s ecosystem, including its CUDA software platform, which has become deeply embedded in AI development workflows.
Industry analysts say the return of H200 shipments, even in a modified form, would provide temporary relief to Chinese customers facing shortages of high-end AI chips. However, they caution that ongoing regulatory uncertainty could limit long-term planning and investment, both for suppliers and buyers.
For Nvidia, the move underscores its strategy of navigating regulatory constraints through product segmentation rather than exiting the Chinese market altogether. The company has repeatedly said it aims to comply fully with U.S. law while continuing to serve customers globally.
If shipments proceed as planned in mid-February, the H200 would become Nvidia’s most advanced AI chip available in China under current rules, reinforcing the country’s access—albeit constrained—to cutting-edge AI infrastructure at a time when global competition in artificial intelligence is intensifying.
