As U.S. restrictions on China's semiconductor industry push Beijing to accelerate self-reliance in chips, a forecast says China's self-sufficiency rate in artificial intelligence (AI) graphics processing units (GPUs) could reach 76% by 2030.
According to Gongshang Shibao on the 25th, Morgan Stanley said China's AI GPU self-sufficiency rate, which was 33% in 2024, could more than double by 2030.
As the United States tightens export controls on advanced AI chips, China is pouring funds into chip localization in response. In addition to established giants such as Huawei and Alibaba, startups including Cambricon and Moore Threads are tightening the reins on developing semiconductors that can replace Nvidia.
It analyzed that the size of China's AI chip market could grow from $6 billion (about 9 trillion won) in 2024 to $51 billion (about 76.1 trillion won) in 2030, with a compound annual growth rate (CAGR) reaching 42%.
China's monthly production capacity for sub-12-nanometer process products was only 8,000 units (based on 12-inch wafers) last year, but it is expected to increase to 20,000 in 2027 and 42,000 in 2028, allowing it to meet domestic demand in key areas. The figure is projected to reach 50,000 in 2030, with the Production yield also expected to rise to 50%.
The report said capital expenditure in China's cloud industry will reach $130 billion (about 194.1 trillion won) in 2030, with about 51% invested in AI GPU-related equipment. It projected the GPU market will grow at an average annual rate of about 23% from 2024 to 2030, expanding to around $67 billion (about 100 trillion won).
Regarding China's semiconductor industry, it also assessed that expansion of production capacity in advanced manufacturing processes, stabilization of the Production yield, and policy support are proving effective.
While High Bandwidth Memory (HBM) and deep ultraviolet (DUV) lithography equipment still depend on overseas companies, some analyses said progress has been made in certain core technologies.
Meanwhile. It noted that while government-led policies can accelerate industrial development, commercial competitiveness is crucial in the medium to long term, adding that Chinese AI GPU companies need to demonstrate the economic impact and profits that would allow them to continue growing beyond 2028.