Nvidia's Share of China's AI Chip Market Falls to 34% as Huawei Closes In at 28%
TECHWORLD · · 1 views
✦ AI Summary
U.S. export restrictions to China and the Chinese government's policy of encouraging domestically produced AI semiconductors have reduced Nvidia's position in China's AI chip market for data centers.
In the first half of 2026, Nvidia's share of the Chinese market was 34%, and shipments fell 44% year on year to 678,000 units.
Over the same period, Huawei shipped 572,000 units to take a 28% share, while shipments of non-GPU AI accelerators reached 1.016 million units, surpassing GPU shipments of 991,000 units.
According to the semiconductor industry on the 16th, U.S. export restrictions to China are limiting Nvidia's sales of cutting-edge AI chips in China. At the same time, the Chinese government is encouraging data center and cloud operators to adopt domestically produced AI semiconductors. IDC said Nvidia's weakness in the Chinese market stems from tighter U.S. export controls and Beijing's policy of expanding the adoption of homegrown semiconductors.
As a result, Nvidia's share of China's AI chip market for data centers in the first half of 2026 stood at 34%. Nvidia's shipments came to 678,000 units. That was down 44% from a year earlier.
Nvidia has seen its position in China's AI chip market for data centers shrink due to the impact of U.S. export restrictions on China and the Chinese government's push for semiconductor self-reliance. In contrast, Chinese chipmakers, including Huawei, expanded shipments. As a result, Chinese semiconductor companies also increased their market share.
IDC cited China's policy of encouraging domestically produced AI semiconductors as the backdrop for the expansion of local companies' presence. Against this trend, the competitive landscape in China's AI chip market is shifting from a GPU-centric structure to one that makes use of a range of AI accelerators.
In China, Chinese companies now need regulatory approval to import Nvidia products, which is cited as another factor in the changing market environment.
Meanwhile, Nvidia CEO Jensen Huang has repeatedly criticized U.S. export restrictions, saying they have handed Huawei a roughly USD 50 billion market in China's AI chip market, equivalent to about KRW 67.189 trillion.
Huawei is actually catching up to Nvidia by expanding shipments. Huawei's AI chip shipments in the first half of 2026 reached 572,000 units, up from more than 300,000 units a year earlier. Huawei's market share was 28%.
Other companies in China are also moving to expand their presence. Alibaba's semiconductor unit T-Head held a 7% share, Cambricon 5%, Baidu's Kunlunxin 4%, and Hygon 3%.
It was analyzed that U.S. export restrictions and the Chinese government's localization policy are simultaneously expanding China's AI semiconductor supply base. IDC said China's local AI semiconductor ecosystem is growing, and noted that shipments rose for 9 of the 11 Chinese semiconductor companies it tracks.
Against this backdrop, changes are also emerging in the GPU-centric structure that had been the main axis of China's AI chip market. As adoption of dedicated AI accelerators optimized for specific AI tasks rapidly expands, the composition of AI chips used in data centers is also changing.
Total shipments of AI chips for data centers in China in the first half of 2026 were 2.008 million units. Of that, GPU shipments were 991,000 units, down 26% from a year earlier.
By contrast, shipments of non-GPU AI accelerators in the first half of 2026 totaled 1.016 million units. Shipments of non-GPU AI accelerators rose by about 80%, and as a result, AI accelerator shipments overtook GPU shipments.
Huawei, Alibaba, Baidu, and Cambricon accounted for 95% of total shipments in the segment. Adoption of products from these four companies is expanding, mainly among government agencies, internet companies, and telecommunications operators.
In the Chinese market, there is a shift in which supply stability and localization requirements are becoming more important in addition to chip computing performance. Accordingly, the trend is forming in which chip performance alone is not enough and supply stability and localization requirements are also taken into account.
In this environment, tailored accelerators can be used for specific AI tasks, and companies can also pursue system configurations tailored to their workloads. The use of customized accelerators is cited as a driver of wider adoption.
According to IDC analysis, the center of competition in AI semiconductors is shifting from individual chip performance to the interconnect performance of entire AI chip clusters and system-level capabilities. The focus of competition is also moving from chip performance to system optimization, and beyond chip computing performance, overall system efficiency, including software and networks, is emerging as a factor that determines competitiveness.
IDC forecast that total annual AI accelerator shipments in 2026 will rise 19% from a year earlier. By contrast, annual GPU shipments in 2026 are expected to fall 9%. In the first half, AI accelerator shipments overtook GPU shipments, and this reversal in GPU and non-GPU accelerator shipments is seen as likely to continue in the annual market as well.
An industry expert said changes in China's AI semiconductor market have implications for Nvidia that go beyond a simple decline in market share. He said that, as U.S. export restrictions have created supply constraints for Nvidia products, Chinese companies are expanding their ability to supply products and build systems based on their domestic market.
Source: TECHWORLD · Park Gyu-chan
Original: https://www.epnc.co.kr/news/articleView.html?idxno=408305
References
This article was produced with the help of an automated content generation algorithm.
Source: TECHWORLD
View originalThis article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.