Nvidia Concedes China AI Chip Market to Huawei

Nvidia CEO Jensen Huang stated that the company has "largely conceded" China's artificial intelligence chip market to domestic rival Huawei, citing the impact of U.S. export restrictions [1]. This admission comes despite Nvidia reporting an 85% surge in quarterly revenue to $81.62 billion, as the company acknowledges it has "evacuated" the region amid shifting geopolitical dynamics [1]. Huang noted that Huawei is "very, very strong" and benefiting from a local ecosystem of chip companies [1].

Key takeaways

• Nvidia CEO Jensen Huang says the company has "largely conceded" the Chinese AI chip market to Huawei [1].
• Nvidia's official market share in China has dropped to zero from roughly 95% prior to U.S. export curbs [2].
• Huawei claims its chips will achieve transistor density equivalent to 1.4nm processes within five years [3].
• Beijing has reportedly refused to approve purchases of Nvidia's H200 chips, steering firms toward domestic suppliers [2].

Strategic retreat from a major market

While Nvidia delivered what sources described as a blockbuster quarter, the company is effectively removing China from its future financial guidance [2]. Huang told investors that he expects no data center compute revenue from the region, a significant shift for a company that previously relied on China for about one-fifth of its data center business [2]. Nvidia disclosed in a recent regulatory filing that it is "effectively foreclosed from competing in China's data center computing market" [2]. Although the U.S. Commerce Department reportedly approved licenses for major Chinese firms to purchase up to 75,000 H200 chips each, Beijing has not approved a single purchase, with reports indicating the government is steering domestic companies away from foreign technology [2].

Huawei’s technological advances

As Nvidia exits the market, Huawei is advancing its own semiconductor capabilities to mitigate the impact of U.S. sanctions that restrict access to advanced lithography tools [3]. The company announced a new principle called the "Tau Scaling Law," which focuses on reducing signal travel time within chips to improve performance and density, aiming for transistor density equivalent to 1.4nm processes in five years [3]. Huawei's Ascend chips are increasingly central to Chinese AI infrastructure, powering models like DeepSeek’s V4, and demand for its flagship Ascend 950PR has reportedly driven prices up by 20% [2][3].

Why it matters

The loss of the Chinese market represents a major restructuring of the global AI semiconductor landscape, with U.S. export controls successfully limiting Nvidia's access while accelerating China's domestic development [1][2]. Chinese firms, including ByteDance, are increasing capital expenditures toward domestic chipmakers, signaling a long-term decoupling of AI supply chains [2]. Huawei claims its new scaling principles will allow it to reach the global frontier of chip manufacturing by the end of the decade, despite current equipment restrictions [3].