Nvidia is publicly stating that the rapid advancement of Chinese AI models benefits the global market and is preparing to sell more advanced chips to China.

This shift in positioning comes as the company attempts to navigate a volatile landscape of U.S. export controls and increasing competition from domestic Chinese hardware. Because Nvidia relies on high volume for its growth, the company views the expansion of the AI ecosystem, regardless of the region, as a catalyst for hardware demand.

Jensen Huang said that better AI models drive broader adoption of the technology, which in turn increases the need for the massive data centers that power them. According to the company, this cycle creates more opportunities for Nvidia to sell its GPUs and networking infrastructure [1], [3].

However, the company faces significant regulatory headwinds. On May 31, the U.S. Department of Commerce moved to close a year-old loophole that had allowed some Nvidia AI chip shipments to Chinese firms operating outside of China [4]. This regulatory tightening has contributed to a surge in the grey market, where prices for Nvidia AI servers and GPUs in China have more than doubled [5].

Despite these restrictions, Nvidia is reportedly telling Chinese customers it could be ready to sell more advanced AI chips soon [2]. This suggests a strategy of maintaining customer relationships while waiting for potential policy shifts or the development of compliant, high-performance hardware.

At the same time, China is reducing its reliance on American technology. One Chinese lab, Z.AI, recently built a one-gigawatt data center using only Chinese-made chips [6]. This development highlights a growing contradiction: while Nvidia believes Chinese AI progress will fuel its own sales, Chinese firms are proving they can scale infrastructure without Nvidia hardware [1], [6].

Better AI models, regardless of where they’re developed, will drive more AI adoption and more demand for Nvidia’s hardware.

Nvidia is attempting to decouple the success of Chinese AI software from the restrictions on its hardware. By framing Chinese AI progress as a global win, the company maintains a narrative of growth and demand. However, the emergence of massive, non-Nvidia data centers in China suggests that the company may be losing its grip on a critical market, making its ability to navigate U.S. export laws a matter of long-term survival in the region.