The United States’ efforts to restrict the export of advanced artificial intelligence (AI) chips to China may create opportunities for Huawei Technologies to expand in its $7 billion domestic market, as analysts suggest these restrictions could lead Nvidia to step back from the Chinese market.
Historically, Nvidia has held a dominant position as the leading provider of AI chips in China, boasting a market share of over 90%.
However, Chinese companies like Huawei have been working on developing their own versions of Nvidia’s popular chips, including the A100 and H100 graphics processing units (GPU).
Huawei’s Ascend AI chips have been compared to Nvidia’s in terms of raw computing power, though they still lag behind in terms of performance.
The restrictions imposed by the U.S. may alter the landscape for Chinese firms. Jiang Yifan, chief market analyst at Guotai Junan Securities, believes that the U.S. move is a significant advantage for Huawei’s Ascend chips. However, there are several challenges ahead.
Many cutting-edge AI projects rely on CUDA, a programming architecture pioneered by Nvidia, which has led to a massive global ecosystem capable of training highly advanced AI models.
Huawei’s alternative, called CANN, is considered more limited in terms of AI model training capabilities, making it less of a direct replacement for Nvidia.
To compete with Nvidia in the Chinese market, Huawei must replicate Nvidia’s ecosystem, including supporting clients in migrating their data and models to Huawei’s platform.
Additionally, intellectual property rights pose a challenge, as many key patents for GPUs are held by U.S. firms.
If Huawei manages to capture Nvidia’s market share, it could score another victory against U.S. export controls, which have targeted the company since 2019.
Over the past year, Huawei has demonstrated resilience by unveiling advanced chips, including smartphone chips, and focusing on becoming a key provider of computing power for AI.
Huawei’s partnerships in China, including iFlyTek and state-owned software firms Tsinghua Tongfang and Digital China, highlight its push to support the development of large language models (LLMs) with homegrown hardware and software technology.
Despite Nvidia’s ecosystem dominance, analysts suggest that domestic players in China, given sufficient time and a substantial customer base, can potentially overcome these challenges.
China’s self-sufficiency drive, championed by President Xi Jinping, is expected to bolster this effort, with a small disruption to near-term supplies potentially leading to a significant boost in long-term self-sufficiency goals.