China’s Tech Giants Expect Domestic Suppliers to Meet Just 20% of AI Chip Demand
Major Chinese technology companies are struggling to secure enough locally manufactured AI processors to support their expanding artificial intelligence operations, highlighting the country's continued dependence on foreign semiconductor technology.
China’s biggest technology companies are facing a significant shortage of domestically manufactured artificial intelligence chips, with local suppliers expected to meet only a fraction of their growing requirements. Companies including Alibaba, Tencent and Baidu are expanding their AI infrastructure, but domestic semiconductor manufacturers are struggling to produce enough advanced processors to satisfy demand. The situation highlights the difficulties Beijing faces in achieving its ambition of building an independent semiconductor industry.
The shortage comes as China accelerates efforts to reduce its dependence on American chipmaker Nvidia following years of US export restrictions. Domestic manufacturers such as Huawei, Cambricon and Alibaba’s T-Head are developing alternative AI processors, but production capacity and technological limitations remain significant obstacles. Advanced memory components, particularly high-bandwidth memory (HBM), are also creating bottlenecks for Chinese chip production.
Despite government support and substantial investments in local semiconductor manufacturing, Chinese technology companies continue to seek alternative sources of computing power. Some businesses have explored purchasing less advanced Nvidia processors, while others are considering leasing computing capacity from overseas data centres equipped with powerful American chips. However, these arrangements face increasing scrutiny from Washington and Beijing as both governments tighten their technology policies.
The supply shortage could complicate China’s ambitions to compete with the United States in developing advanced AI models and infrastructure. Although domestic chipmakers are making progress, challenges involving manufacturing technology, memory supply and software compatibility remain difficult to overcome. The situation also demonstrates that building a competitive semiconductor industry requires more than financial investment, as China must develop a complete ecosystem capable of supporting large-scale AI development.
