China faces significant technical and commercial barriers in developing domestically-designed AI chips to rival Nvidia.
Chinese companies are accelerating efforts to develop viable alternatives to Nvidia's AI chips as Beijing seeks to reduce dependence on American technology, facing significant headwinds from U.S. sanctions and Nvidia's market dominance. Nvidia's dominance has been driven by large cloud computing players purchasing its server products containing graphics processing units, or GPUs, which enable companies such as OpenAI to train massive AI models. The U.S. government has restricted exports of Nvidia's most advanced chips to China since 2022, with restrictions tightening further last year—restrictions that are key to China's ambitions to become a leading AI player.
China's leading contenders seeking to challenge Nvidia include technology giants Huawei, Alibaba and Baidu, along with startups such as Biren Technology and Enflame. According to Wei Sun, a senior analyst at Counterpoint Research, "These companies have made notable progress in developing AI chips tailored to specific applications (ASICs). However, competing with Nvidia still presents substantial challenges in technological gaps, especially in general-purpose GPU. Matching Nvidia in short-term is unlikely." Chinese firms face a "lack of technology expertise," Sun noted, representing one of several challenges to their efforts.
The greatest obstacles stem from U.S. sanctions and their ripple effects. Many of China's leading Nvidia challengers have been placed on the U.S. Entity List, restricting access to American technology, while various U.S. curbs have prevented key AI-related semiconductors and machinery from being exported to China. China's GPU players design chips but rely on manufacturing partners, historically Taiwan Semiconductor Manufacturing Co., or TSMC; however, U.S. restrictions now prevent many from accessing TSMC's production. These companies must instead turn to SMIC, China's largest chipmaker, whose technology lags TSMC by generations partly because Washington has restricted SMIC's access to advanced machinery from Dutch firm ASML required for manufacturing cutting-edge chips. Additionally, Huawei's development of advanced smartphone and AI chips is consuming significant SMIC capacity, as Paul Triolo, a partner at Albright Stonebridge, explained: "The key bottleneck will be domestic foundry leader SMIC, which will have a complex problem of dividing limited resources for its advanced node production between Huawei, which is taking up the lion's share currently, the GPU startups, and many other Chinese design firms which have been or may be cutoff from using global foundry leader TSMC to manufacture their advanced designs."
Beyond hardware, Nvidia's ecosystem advantage—particularly its CUDA software platform allowing developers to create applications for its hardware—represents a formidable barrier. "This is the key, it is not just about the hardware, but about the overall ecosystem, tools for developers, and the ability to continue to evolve this ecosystem going forward as the technology advances," Triolo said. Huawei has emerged as a leading contender with its Ascend series of data center processors, currently offering the Ascend 910B with the upcoming Ascend 910C potentially matching Nvidia's H100 product. While Huawei "holds lots of advantages" in software and developer community building, it faces similar challenges as competitors: "The GPU software support ecosystem is much more entrenched around Nvidia and to a lesser degree AMD, and Huawei faces major challenges, both in producing sufficient quantities of advanced GPUs such as part of the Ascend 910C, and continuing to innovate and improve the performance of the hardware, given U.S. export controls that are limiting the ability of SMIC to produce advanced semiconductors," Triolo stated.
Recent setbacks underscore the difficulty of the undertaking. In 2022, Biren Technology conducted layoffs, followed by Moore Threads the following year, both blaming U.S. sanctions. However, startups remain determined, with Bloomberg reporting last week that both Enflame and Biren are seeking public listings to raise capital. According to Triolo, "Biren and the other GPU startups are staffed with experienced industry personnel from Nvidia, AMD, and other leading western semiconductor companies, but they have the additional challenge of lacking the financial depth that Huawei has. Hence both Biren and Enflame are seeking IPOs in Hong Kong, to raise funding for additional hiring and expansion."