China data centres are shifting to local chips as Beijing steps up efforts to reduce dependence on foreign technology amid tightening US export curbs. The move highlights the country’s push for tech autonomy while supporting its rapidly growing AI sector.
Publicly owned computing hubs across China have been instructed to source more than half of their chips from domestic suppliers, according to South China Morning Post. The directive is part of Beijing’s drive to strengthen its semiconductor sector and cut reliance on foreign technology.
The policy stems from guidelines first introduced in March 2023 by Shanghai authorities, which required the city’s intelligent computing centres to adopt over 50 percent local chips by 2025. Those rules have now been expanded into a nationwide mandate backed by the Ministry of Industry and Information Technology (MIIT) and the National Development and Reform Commission (NDRC).
China’s push comes as Washington continues to tighten export restrictions, blocking access to advanced semiconductors such as Nvidia’s H100 and H800. While Chinese chips are still considered less powerful, they are increasingly being used in inferencing tasks for AI models, though Nvidia remains dominant in training new systems.
The country has also accelerated the rollout of intelligent computing centres, with more than 500 new projects launched in 2023 and 2024 in regions including Inner Mongolia and Guangdong, according to KZ Consulting. These facilities are designed to meet the growing computing demands of China’s AI industry.
Although some firms like iFlytek have turned to Huawei chips for AI training, experts say it remains unclear how quickly China’s domestic offerings can fully replace US technology in the data centre space.