The Rise of Cambricon Technologies in AI Chip Development
Cambricon Technologies is a pivotal player in China’s ambitious endeavor to rival the United States in artificial intelligence (AI) leadership. While less prominent than major tech giants like Huawei and Moore Threads, Cambricon specializes in AI-focused Graphics Processing Units (GPUs) that showcase significant growth potential. In August 2025, the company received approval from the Shanghai Stock Exchange to raise an impressive $560 million for the design of four new chips aimed at enhancing AI model training and inference, alongside developing an alternative to NVIDIA’s CUDA.
According to South China Morning Post, Cambricon announced a remarkable net profit of 2,060 million yuan (around 257 million euros) for the fiscal year 2025. While this may seem modest in the semiconductor and AI landscape, it reflects substantial growth for a company founded only a decade ago, especially considering a 450% increase in income compared to the previous year.
Competition and Growth in the Chinese AI Chip Market
Cambricon is not the sole representative of China’s semiconductor capabilities. Competitors like Moore Threads recently reported exceptional economic performance, bolstered by increasing domestic demand for semiconductors as the Beijing government strives for technological autonomy.
China’s Resource Control in Semiconductor Production
The semiconductor manufacturing landscape is multifaceted, with China controlling a significant portion of the supply chain. Producing approximately 70% of the world’s rare earth elements and controlling 90% of the processing industry, China is a formidable player in this arena. Notably, it refines 99% of the heavy rare earths globally, underpinning its strength in high-tech industries such as electronics and renewable energy.
Key Players in AI GPU Development
In the competitive GPU market for AI applications, several Chinese firms have emerged as contenders. Cambricon’s flagship products, including the MLU series (Machine Learning Unit) and Siyuan models, are designed to compete against offerings from NVIDIA and Huawei. Meanwhile, Moore Threads has developed GPUs such as the MTT S4000 and MTT S3000, aiming to match the performance of NVIDIA and AMD products.
The Challenge of Advanced Semiconductor Manufacturing
Huawei is another critical player in the Chinese AI chipset industry, with its Ascend 910D and Ascend 920 GPUs gaining traction among local businesses engaged in AI development. A pressing challenge for China is to establish its advanced semiconductor manufacturing capabilities; without 100% Chinese-made high-performance chips, the nation risks jeopardizing its military capabilities, AI model development, and overall technological competitiveness.
Government Support and Future Prospects
To enhance its semiconductor manufacturing sector, the Chinese government has begun supporting firms capable of developing advanced photolithography technology. Companies like SiCarrier, Shanghai Yuliangsheng, and Shanghai Micro Electronics Equipment (SMEE) benefit from substantial subsidies aimed at fostering innovation and self-reliance.
As the competition intensifies, time is of the essence for China to secure its standing in the global technology arena. The ongoing technological rivalry between China and the U.S. primarily revolves around control of advanced manufacturing technologies, especially in semiconductors.
For more insights into the dynamics of the Chinese AI chip industry, visit the South China Morning Post.