Nvidia’s grip on China’s data-centre AI chip market slipped to 34% in the first half of 2026, according to an IDC report seen by Reuters. The US chipmaker shipped 678,000 AI accelerators in China from January through June, down 44% from a year earlier. This decline tracks alongside US export controls that limit Nvidia’s ability to sell its most powerful AI chips in the country.
Washington has restricted China’s access to Nvidia’s advanced AI chips, including models such as the B200 and B300, which are barred from export. Despite these restrictions, some of these cutting-edge chips have reportedly been smuggled into China. Meanwhile, Beijing has pushed domestic alternatives and mandated that Chinese companies obtain approval to import Nvidia’s compliant chips.
Nvidia CEO Jensen Huang criticised these export curbs repeatedly, arguing that they forced the company to surrender China’s $50 billion AI chip market to Huawei. IDC did not specify which Nvidia products contributed to the 34% market share figure or confirm the inclusion of the smuggled chips.
Huawei advanced significantly in the same period, shipping 572,000 AI accelerators, rising from over 300,000 a year earlier, and claiming 28% of the overall market. This surge reflects Beijing’s broader efforts to boost local suppliers amid the US restrictions.
Other Chinese players also made inroads. Alibaba’s chip-design unit T-Head held 7%, Cambricon 5%, Baidu’s Kunlunxin 4%, and Hygon 3%. IDC tracked 11 Chinese chipmakers in total, with nine increasing shipments compared to the previous year.
Total AI-accelerator shipments in China reached 2.008 million units in the first half of 2026. Shipments of graphics processing units, or GPUs, which are the chips most commonly used to train AI models, fell 26% to 991,000 units. This drop reflects ongoing constraints on GPU supply in mainland China.
Conversely, shipments of other AI accelerators rose nearly 80% to 1.016 million units, surpassing GPUs by volume. This category includes chips designed for specific AI tasks such as running trained models, rather than training them.
Huawei, T-Head, Kunlunxin, and Cambricon combined accounted for 95% of the non-GPU segment, which IDC said was gaining adoption among government agencies, internet companies, and telecom providers.
IDC noted a shift in competition from the performance of individual chips to system-level capabilities, including how well chips function within larger computing clusters.
The report forecasts that total accelerator shipments in China will grow 19% in 2026, even as GPU shipments are expected to decline by 9%.
Nvidia did not respond to requests for comment on the report.
According to Joy Online.
