Nvidia Loses Ground in China AI Chip Market as Huawei Captures 28% Share

Nvidia's China AI-Chip Share Falls to 34% Amid Export Restrictions

Nvidia's grip on the China AI chip market weakened significantly in the first half of 2026, with the company's market share dropping to 34% according to an IDC report seen by Reuters1

1

. The US chipmaker shipped 678,000 AI accelerators between January and June, representing a sharp 44% decline from the previous year2

2

. This contraction stems directly from Washington's tightened US export controls, which restrict China's access to Nvidia's most powerful AI chips and limit the range of products the company can sell in the world's second-largest economy1

1

. Beijing has simultaneously promoted domestic alternatives and required Chinese companies to seek approval before importing even compliant Nvidia chips, creating additional barriers for the American semiconductor giant2

2

.

Huawei and Domestic Rivals Gain Ground in Data-Center AI Chip Market

While Nvidia stumbled, Huawei surged forward by shipping 572,000 AI accelerators during the same period, up from more than 300,000 units a year earlier1

1

. This performance secured Huawei a 28% share of the overall data-center AI chip market, positioning the Chinese technology leader as Nvidia's closest competitor3

3

. The competitive landscape now features multiple domestic rivals making significant inroads. Alibaba T-Head captured 7% market share, followed by Cambricon with 5%, Baidu Kunlunxin with 4%, and Hygon with 3%1

1

. Nine of the 11 Chinese chipmakers tracked by the IDC report increased shipments year-over-year, demonstrating the breadth of China's accelerating semiconductor ecosystem2

2

.

Total AI Accelerator Shipments Reach 2.008 Million Units

Despite Nvidia's decline, China's overall demand for AI accelerators remained robust. Total AI-accelerator shipments reached 2.008 million units in the first half of 20261

1

. The composition of these shipments reveals a fundamental shift in the market. Graphics processing units, or GPUs—the chips most widely used to train advanced AI models—fell 26% to 991,000 units as supplies eligible for sale in mainland China remained limited2

2

. Meanwhile, shipments of non-GPU AI chips rose nearly 80% to 1.016 million units, overtaking GPUs by volume for the first time1

1

. This category includes chips designed for specific AI tasks such as running trained models rather than training them from scratch.

Geopolitical Tensions Reshape Market Dynamics

The erosion of Nvidia's position reflects intensifying geopolitical tensions between Washington and Beijing over semiconductor technology. Nvidia CEO Jensen Huang has repeatedly criticized Washington's export curbs, stating the policy forced a US company to cede China's $50 billion AI chip market to Huawei1

1

. The IDC report did not specify which Nvidia products comprised its 34% market share or whether the figure included cutting-edge chips such as the B200 and B300, which are barred from export to China but have reportedly been smuggled into the country2

2

. Huang acknowledged in May that Huawei was "very, very strong" and admitted Nvidia had "largely conceded" the China market as export controls accelerated development of the domestic chip ecosystem3

3

.

Competition Shifts to System-Level Capabilities

The China AI chip market is evolving beyond individual chip performance. According to the IDC report, competition is shifting from the performance of a single chip to system-level capabilities, including how well chips work in larger computing clusters1

1

. Huawei, T-Head, Kunlunxin, and Cambricon together accounted for 95% of the non-GPU segment, which is being adopted by government agencies, internet companies, and telecoms providers2

2

. This adoption pattern suggests Chinese organizations are increasingly comfortable deploying domestically-produced AI accelerators for critical infrastructure. Huawei's hardware expansion is being reinforced by strategic software partnerships, including a collaboration with DeepSeek on open-source software for Ascend chips that directly challenges Nvidia's CUDA platform3

3

.

Market Outlook: Diverging Trajectories for GPU and Non-GPU Segments

Looking ahead, IDC expects total accelerator shipments in China to grow 19% in 2026, even as GPU shipments fall 9%1

1

. This forecast underscores the diverging trajectories of different chip categories. The continued decline in GPU availability will likely accelerate adoption of alternative AI accelerators designed by domestic rivals, further entrenching local suppliers in China's AI infrastructure. For Nvidia, the narrowing gap with Huawei raises questions about whether the company can maintain its leadership position in a market where regulatory barriers and nationalistic procurement policies increasingly favor homegrown solutions. The competitive dynamics suggest US export restrictions, rather than protecting American technological advantage, may be catalyzing the very outcome they sought to prevent: a self-sufficient Chinese semiconductor industry capable of supporting advanced AI models without reliance on American technology.