Nvidia Concedes China AI Chip Market to Huawei

In a candid acknowledgment of shifting dynamics in the global AI chip landscape, Nvidia CEO Jensen Huang stated that the company has "largely conceded" China's AI chip market to domestic rival Huawei. The remarks came during Nvidia's earnings call on May 21, 2026, as the company reported yet another blockbuster quarter, driven by surging demand for its data center GPUs outside of China.

Market Impact: A Strategic Retreat

Huang's concession underscores the profound impact of U.S. export controls, which have severely restricted Nvidia's ability to sell its advanced AI chips—such as the A100, H100, and subsequent models—to Chinese customers. Since the initial restrictions in 2022, China's tech giants have increasingly turned to homegrown alternatives, with Huawei emerging as the primary beneficiary. Huawei's Ascend 910B and newer 920 series chips now power a significant portion of China's AI infrastructure, displacing Nvidia in key data centers and research institutions.

According to industry analysts, Huawei's market share in China's AI chip segment has surged from less than 10% in 2022 to over 50% in early 2026, while Nvidia's share has plummeted from nearly 90% to around 30%. The remaining share is split among other domestic players like Cambricon and Biren Technology.

Regulatory and Strategic Dimensions

Huang's comments highlight a broader strategic dilemma for Nvidia. While the company continues to dominate markets outside China, the loss of the world's second-largest economy represents a significant long-term risk. China's domestic AI chip ecosystem is rapidly maturing, supported by government policies and substantial investments. Huawei, in particular, has leveraged its deep expertise in telecommunications and chip design to create competitive AI accelerators that, while not matching Nvidia's absolute performance, offer sufficient capability for many workloads.

Despite the concession, Huang emphasized that Nvidia remains committed to serving Chinese customers and is actively seeking approval from the U.S. government to export modified chips that comply with regulations. However, he expressed low expectations for approval, citing ongoing geopolitical tensions. "We want to return to China, but the path forward is uncertain," Huang stated.

Adoption and Ecosystem Considerations

One critical factor in Nvidia's retreat is the software ecosystem. Nvidia's CUDA platform has long been a moat, but Huawei's MindSpore and other domestic frameworks are gaining traction, reducing the switching costs for Chinese developers. Additionally, Chinese hyperscalers like Alibaba, Baidu, and Tencent have invested heavily in optimizing their AI stacks for Huawei's architecture, further entrenching the shift.

From a technical perspective, Huawei's chips are now competitive in training and inference for many AI models, particularly those optimized for Chinese language and applications. While Nvidia still leads in cutting-edge performance, the gap is narrowing, and for many Chinese enterprises, the combination of adequate performance, supply security, and government preference makes Huawei the default choice.

1. Nvidia has largely conceded China's AI chip market to Huawei due to U.S. export controls.
2. Huawei's market share in China's AI chip market has surpassed 50%, while Nvidia's has dropped to around 30%.
3. Nvidia remains interested in returning to China but faces low approval prospects for chip exports.
4. The shift is reinforced by China's growing domestic software ecosystem and government support for local alternatives.
5. Despite the loss, Nvidia continues to report strong global growth driven by demand outside China.

Sources: CNBC