Beijing's Directive on AI Chips
China has issued a new directive mandating that all state-funded data centers must exclusively utilize domestically produced artificial intelligence (AI) chips. This move signals a significant escalation in Beijing's push for technological self-reliance and aims to reduce dependence on foreign, particularly U.S., semiconductors. The guidelines, reported around November 5-7, 2025, apply not only to new data center projects but also retroactively to those with less than 30% completion progress, requiring the replacement of any foreign AI chips already installed.
This policy marks a formal shift from earlier measures that merely discouraged foreign chip purchases. Previously, requirements suggested at least 50% domestic AI chips, but this has now been strengthened to 100% for state-funded projects.
Impact on Foreign and Domestic Chipmakers
The directive is expected to significantly impact major U.S. semiconductor firms, including Nvidia, AMD, and Intel, which have traditionally supplied chips to Chinese data centers. Nvidia, a dominant player with over 90% of the global AI chip market, is particularly affected. Even its H20 chips, specifically designed to comply with U.S. export controls, are now reportedly off-limits. Nvidia CEO Jensen Huang has stated that the company is 'effectively blocked' from selling its AI chips into China for now, with 'no active discussions' to ship products there.
Conversely, the ban is anticipated to create substantial opportunities for Chinese domestic chipmakers. Companies such as Huawei (with its Ascend line), Cambricon, Enflame, MetaX, and Moore Threads are poised to benefit from the increased demand for homegrown alternatives.
Strategic Push for Technological Self-Reliance
This aggressive policy is a direct response to ongoing U.S.-China tech tensions and export controls, which Beijing views as threats to its technological advancement. China sees AI capabilities as critical for both national and economic security, making 'independent and controllable' AI a key objective. The move aligns with China's broader strategy for technological independence, as outlined in its 15th five-year plan (2026-2030), which prioritizes self-reliance in semiconductors and artificial intelligence.
Since 2021, Chinese AI data center projects have attracted over $100 billion in state funding, underscoring the government's commitment to building robust domestic AI infrastructure. While Chinese chipmakers have made strides, their performance and software tooling still lag behind U.S. counterparts. However, China's abundant low-cost electricity may help offset some of the efficiency gaps in domestic chips.
Broader Implications for U.S.-China Tech Rivalry
The directive marks one of Beijing's strongest steps yet to eliminate foreign technology from critical infrastructure, intensifying the competition for dominance in high-performance computing and artificial intelligence. This follows previous measures, including a 2023 ban on Micron Technology products in critical infrastructure, which led to the U.S. memory chipmaker's exit from the Chinese server chip market. The latest ban underscores China's determination to reduce its reliance on Western technology amid growing geopolitical uncertainty.
5 Comments
ZmeeLove
A bold and decisive step. This shows real commitment to their long-term vision.
Habibi
So much for global collaboration. This will only widen the tech divide.
Bermudez
Investing in their own chipmakers is the only way forward. Great strategy!
ZmeeLove
The move demonstrates China's firm commitment to its long-term strategic goals for AI dominance. But, achieving this through a complete ban on superior foreign tech might lead to internal innovation struggles until their domestic firms truly catch up.
Bermudez
It's an understandable reaction to ongoing US tech restrictions, aiming to future-proof their critical infrastructure. Yet, the immediate impact on cost and efficiency, given the current state of domestic chip technology, could be substantial.