China Proposes $295 Billion AI Data Center Network Fueled by Domestic Chips
China is planning a $295 billion, five-year initiative to create a national AI data center grid. The project aims for 80% domestic technology, including chips, by 2028. However, limitations in local chip production, particularly advanced processors and high-bandwidth memory, could hinder the ambitious goal and restrict reliance on foreign suppliers like Nvidia.
Key points
- China is drafting a national plan to invest approximately 2 trillion yuan ($295 billion) over five years in building a nationwide AI data center grid.
- The initiative, led by the National Development and Reform Commission, aims for at least 80% of the technology, including AI chips, to be sourced from domestic suppliers.
- State carriers China Mobile and China Telecom are slated to operate the facilities, forming a unified computing grid by 2028.
- The plan faces potential limitations due to China's domestic chip manufacturing capacity, especially concerning advanced nodes and high-bandwidth memory.
- This constraint could restrict the use of foreign accelerators like Nvidia and AMD, impacting the realization of the 80% domestic sourcing target.
China is reportedly planning a significant national infrastructure project centered around artificial intelligence. Sources indicate the government is drafting a five-year plan that would allocate roughly 2 trillion yuan, or $295 billion, towards establishing a nationwide network of AI data centers.
The ambitious initiative aims to foster technological self-reliance, with a stated goal of sourcing at least 80% of the underlying technology, including AI-specific chips, from domestic Chinese manufacturers. Companies like Huawei are expected to play a key role in supplying this technology. The National Development and Reform Commission is reportedly overseeing the blueprint for this network, with state-owned telecommunications giants China Mobile and China Telecom set to manage the majority of the facilities and interconnect them into a single computing grid by the year 2028.
However, the plan faces considerable challenges related to China's domestic semiconductor industry. Experts point to limitations in local chip production, particularly in advanced manufacturing processes and the supply of high-bandwidth memory (HBM), which is crucial for high-performance AI computing. The country's most advanced foundries are operating at near-full capacity, and a strong reliance on domestic components could effectively exclude leading foreign chipmakers like Nvidia and AMD, potentially capping the grid's performance and development based on available local production.
Sources
The WireByte editorial team synthesises technology news from multiple primary sources, verifies the facts, and links every source. Articles are produced with AI assistance and reviewed under our editorial policy.