Nvidia’s efforts to resume large-scale AI chip sales in China remain stalled despite the U.S. government approving export licenses for around 10 Chinese companies, according to the report from Reuters. No deliveries of the H200 chips have taken place so far, leaving billions of dollars in potential business in limbo as Nvidia CEO Jensen Huang travels to Beijing alongside President Donald Trump for talks with Chinese President Xi Jinping.
The U.S. Commerce Department has reportedly cleared companies including Alibaba, Tencent, ByteDance, and JD.com to purchase Nvidia’s H200 chips under export licensing rules. Approved distributors such as Lenovo and Foxconn are also authorized to handle sales. Each approved customer can reportedly buy up to 75,000 chips under the licensing framework.
Despite those approvals, Chinese buyers have delayed purchases after receiving signals from Beijing to proceed cautiously. Sources said Chinese regulators are increasingly concerned that dependence on imported Nvidia hardware could undermine domestic semiconductor development efforts, particularly as Chinese companies promote locally developed AI chips from firms such as Huawei.
The H200 is Nvidia’s second-most powerful AI accelerator and remains highly sought after for training and running advanced artificial intelligence models. Before tighter U.S. export restrictions, Nvidia controlled roughly 95% of China’s advanced AI chip market. China previously represented about 13% of Nvidia’s revenue, and CEO Jensen Huang has estimated the country’s AI market could reach $50 billion this year.
The stalled sales also reflect growing complexity in U.S. export policy. January rules require Chinese buyers to demonstrate strict security procedures and guarantee the chips will not support military uses. Nvidia must also certify sufficient inventory exists within the United States before exports can proceed.
Additional complications emerged after the Trump administration negotiated a structure allowing the U.S. government to receive 25% of revenue from approved sales. To comply with legal requirements, the chips would reportedly need to pass through U.S. territory before shipment to China, a condition that has triggered concerns in Beijing over supply chain security and possible vulnerabilities.
The Geopolitical Cost
For Nvidia, the uncertainty threatens one of the world’s largest AI infrastructure markets. Huang has previously warned that U.S. export restrictions are steadily eroding Nvidia’s position in China, where domestic alternatives are gaining traction. Chinese AI firms increasingly emphasize their use of local chips as Beijing pushes for greater technological self-sufficiency.
The situation also reflects a growing divide inside Washington. Some policymakers argue controlled sales help maintain U.S. influence over China’s AI ecosystem, while critics contend any chip exports could narrow America’s lead in artificial intelligence development.
Market Pressures and Policy Shifts
Nvidia’s relationship with China has faced repeated disruption since the U.S. began tightening semiconductor export controls in recent years. Washington has progressively restricted access to the company’s most advanced AI processors over concerns they could support Chinese military or strategic AI development.
At the same time, Beijing has intensified efforts to reduce reliance on foreign technology suppliers. Recent Chinese supply chain security regulations reportedly triggered a wider government review aimed at limiting foreign dependencies in critical infrastructure.
The dispute now places Nvidia in a difficult position between two competing national strategies. While the company seeks to preserve access to China’s massive AI market, both Washington and Beijing are increasingly treating advanced semiconductors as strategic assets rather than ordinary commercial products.
