A Strategic Push for Chip Sovereignty
According to Reuters, Chinese authorities have directed that all data centres receiving any form of state funding must exclusively use domestically-produced AI chips. For centres still under construction (less than 30% complete), previously installed foreign chips such as those from Nvidia, Intel, and AMD must be removed or procurement plans scrapped. For more advanced projects, decisions will be made on a case-by-case basis.
This sweeping directive, though not officially confirmed by regulators, reportedly stems from a desire to eliminate foreign technological dependency in China's critical infrastructure. It comes amid intensified efforts to build AI self-sufficiency, particularly in high-performance computing, where China has lagged behind U.S. players due to sanctions and export controls.
Implications for Nvidia and the U.S. AI Supply Chain
Nvidia is expected to be the largest casualty. Once dominant in China’s AI chip market with a 95% share in 2022, its presence has dwindled to nearly zero in 2025. The company's H20 chips its most advanced models permitted under U.S. export rules are also included in Beijing’s new restrictions. This move effectively erases any lingering hopes of revenue recovery from China’s $100 billion AI data centre investment drive.
Other companies like Intel and AMD face similar exclusion. Although chips like Nvidia’s B200 and H200 are banned for direct export, they are reportedly accessible in China through grey-market channels. With the new guidance, even this indirect avenue could be closed off for government-affiliated infrastructure.
Strategic Calculus Behind the Ban
The decision aligns with a broader strategy of economic decoupling and self-reliance, especially in light of Washington's semiconductor sanctions, which have hindered China's access to the tools needed for advanced chip fabrication. The U.S. argues these controls are necessary to prevent military misuse of cutting-edge AI chips.
President Trump recently reaffirmed restrictions, stating that China could "deal with Nvidia but not in terms of the most advanced chips." The U.S. position leaves room for lower-tier chip sales but blocks Nvidia's most powerful models, reinforcing Beijing’s motivation to pivot away entirely from U.S. technology.
Domestic Chipmakers: Rising Demand but Facing Limitations
The ban offers a tailwind for Chinese chipmakers such as Huawei, Cambricon, MetaX, Moore Threads, and Enflame. While these firms are actively developing AI hardware alternatives, adoption has been slow due to gaps in performance and ecosystem support especially in software, where Nvidia’s CUDA platform remains an industry standard.
Nonetheless, this policy gives local players a shot at rapid expansion. Already, China has showcased AI data centres powered solely by domestic chips, signaling both ambition and capacity.
Still, the move could create short-term inefficiencies and performance bottlenecks. China’s largest foundry, SMIC, remains constrained by U.S. export controls on chip-making equipment, slowing production of cutting-edge processors.
China’s latest directive deepens the technological bifurcation between the U.S. and China. It signals that even as trade hostilities temporarily ease at the diplomatic level, on the ground, both sides are doubling down on long-term decoupling. For Nvidia and other U.S. firms, this is a significant revenue loss and a potential shift in global chip demand. For China, it is a bet on domestic innovation under pressure and a clear statement that strategic self-reliance now trumps performance or market integration in AI infrastructure.
Source: Reuters
Copyright © 2025 FastBull Ltd
News, historical chart data, and fundamental company data are provided by FastBull Ltd.
Risk Warnings and Disclaimers
You understand and acknowledge that there is a high degree of risk involved in trading. Following any strategies or investment methods may lead to potential losses. The content on the site is provided by our contributors and analysts for information purposes only. You are solely responsible for determining whether any trading assets, securities, strategy, or any other product is suitable for investing based on your own investment objectives and financial situation.