Amidst the ongoing technological trade war between the United States and China, an intriguing development may signal a slight warming of relations. Jensen Huang, CEO of chipmaking giant Nvidia, publicly lauded China’s generative AI models just a day after his company announced it expects to soon resume selling a crucial AI chip to the country. Huang’s statements and Nvidia’s business moves indicate a delicate attempt to navigate a complex relationship, governed on one hand by strict U.S. export restrictions and on the other by significant business interests in the Chinese market.

Praising Chinese AI and Resuming Exports

Jensen Huang spoke on Wednesday at the opening ceremony of a supply chain expo in Beijing, where he praised Chinese models like DeepSeek, Baidu’s Ernie Bots, Alibaba, Tencent, and Minimax, stating they are “world-class, developed here, and openly shared [and] spurred AI developments around the world.” His comments gain further weight given that, according to Huang, “more than 1.5 million developers in China are now building on Nvidia to bring their innovations to life.” These commendations underscore the importance of the Chinese market for Nvidia, both as a customer and as a strategic partner in the global AI development landscape.

These developments come a day after Nvidia announced on Tuesday that it “expects to soon resume shipments of its H20 chips to China,” following assurances from the U.S. government. Shipments of the H20 chips, deemed vital for AI development, were halted in April due to new U.S. requirements at the time. This halt led to a significant financial impact for Nvidia: Huang noted in May that U.S. chip restrictions had “nearly cut” Nvidia’s market share in China. The company also reported a $2.5 billion loss in sales during the April quarter and is likely to absorb an additional $8 billion hit in the July quarter, putting its quarterly sales at $45 billion. The resumption of H20 chip exports could add 10% to Nvidia’s street estimates, highlighting the Chinese market’s importance to the company’s bottom line.

Challenges of US Restrictions and Advantages of Chinese Open Source

Jensen Huang’s current visit is his third to China this year, according to reports from January, indicating his continued efforts to maintain ties with the Chinese market despite the restrictions. U.S. restrictions on advanced semiconductor exports to China stem from concerns that the technology could support Beijing’s defense sector development. Huang has previously countered these concerns, emphasizing that global access is necessary for the U.S. to remain a global leader in AI. He also warned that Chinese companies like Huawei stand to benefit from these restrictions as they develop their own domestic capabilities.

In the global race for AI development, Chinese companies have rushed to release their own AI models, despite limited access to advanced training chips, with even chatbots like ChatGPT not officially available in China. Huang particularly praised the open-source approach adopted by Chinese companies in AI, which allows developers free access to the underlying code. This approach, not yet fully embraced by OpenAI in the U.S., is seen by Huang as a “catalyst for global progress” and a “key” to AI safety, enabling international collaboration on standards. He emphasized how AI “powers” many Chinese consumer technologies, from Tencent’s WeChat to Alibaba’s Taobao shopping app, ByteDance’s Douyin short-video app, and Meituan’s “super convenient” delivery app.

The latest U.S. government restrictions on Nvidia came after a continuous tightening of export controls over the past three years. However, following U.S.-China trade talks held in London last month, there have been signs of some easing: the U.S. has begun to relax some restrictions on high-tech exports to China, while Beijing has simultaneously resumed issuing licenses allowing its companies to export highly needed rare earth metals to the U.S.

Conclusion and Forward Outlook

Jensen Huang’s recent statements and the forecast for resumed H20 chip exports to China reflect the complex dynamics among business interests, government policy, and geopolitical efforts. While the United States aims to limit China’s AI capabilities for national security reasons, companies like Nvidia view the Chinese market as an essential component for their growth and technological leadership. Huang’s praise for Chinese AI and its open-source approach, alongside the noted export relaxations, indicates an attempt to find common ground.

The coming months will reveal whether these developments are a harbinger of some normalization in technology relations between the two powers, or merely isolated relaxations within a prolonged trade war. The impact on the global chip market, AI development, and the overall global economy depends on the delicate balance that will or will not be achieved between security concerns and economic interests.