Telegram is where we really talk. Don't miss out!
In a significant move, the Chinese government has mandated that new data center projects funded by the state must exclusively utilize domestically-produced artificial intelligence chips, according to a report by Reuters.
This directive comes amid rising tensions with the U.S., particularly with former President Trump highlighting four key issues that could escalate trade disputes: Taiwan, democracy and human rights, China’s political system, and development rights.
Chinese regulatory bodies have instructed data centers that are less than 30% complete to eliminate any installed foreign chips or cancel their procurement plans. Projects at more advanced stages will be reviewed on a case-by-case basis.
Some initiatives have already been halted, including a facility in northwestern China that intended to use Nvidia chips, a source revealed. This decision underscores China’s aggressive strategy to reduce reliance on foreign technology for its critical infrastructure and pursue self-sufficiency in AI chip production.
The access to advanced AI chips like those from Nvidia has been a contentious point in U.S.-China relations, as both nations vie for supremacy in high-end computing. Trump mentioned in a recent interview that the U.S. would allow China to manage Nvidia but restrict access to the most advanced chips.
Beijing’s stance is clear: if technology is restricted, they will develop their own.
This policy could hinder Nvidia’s attempts to regain market share in China, while providing local competitors such as Huawei with increased opportunities for chip sales. Other foreign chip manufacturers like AMD and Intel are also affected.
It remains uncertain whether this guidance is applicable nationwide or limited to specific provinces, but it is expected to become a standard practice unless circumstances change. The regulatory bodies responsible for issuing the order have not been disclosed.
Since 2021, state funding exceeding $100 billion has been allocated to AI data center projects in China, as revealed by a Reuters analysis of government contracts. While many data centers have received state support, the exact number affected by the new guidelines is still unclear.
China has long expressed frustration over U.S. export controls aimed at hindering its technological advancement, implementing various retaliatory measures to reduce dependency on American technology. In response, the U.S. has intensified its industrial policy to bolster domestic rare earth production, aiming to lessen reliance on China.
Washington justifies its restrictions by claiming that the Chinese military could exploit these chips to enhance its capabilities. This year, China has also discouraged local tech firms from acquiring advanced Nvidia chips due to security concerns, showcasing a new data center powered entirely by domestic AI chips.
In 2023, Beijing prohibited the use of Micron products in critical infrastructure, leading the largest U.S. memory chipmaker to exit the server chip market in China, according to a Reuters report.
Nvidia CEO Jensen Huang has actively lobbied for increased sales of AI chips to China, arguing that keeping China’s AI industry reliant on U.S. technology serves American interests. The company’s share of the Chinese AI chip market has plummeted to zero from 95% in 2022.
Excluding foreign chipmakers like Nvidia from significant state projects would severely impact their revenue in China, even as a deal is reached to resume advanced chip sales. The new guidelines encompass Nvidia’s H20 chips, the most advanced the company is permitted to sell to China, as well as more powerful processors like the B200 and H200, which are banned from export to China but remain accessible through gray-market channels.
The question now arises: does China possess the technology to achieve parity with the U.S.?
With this latest directive, the Chinese government is further carving out market share for domestic chip manufacturers. China boasts several AI chip companies, with Huawei Technologies leading the pack, alongside smaller firms like Cambricon and startups such as MetaX, Moore Threads, and Enflame.
While products from these domestic companies compete with some of Nvidia’s offerings, they have faced challenges in penetrating the market as developers accustomed to Nvidia’s reliable software ecosystem hesitate to switch to local alternatives.
While this move may bolster sales of domestic chips, it risks widening the gap in AI computing power between the U.S. and China. American tech giants like Microsoft, Meta, and OpenAI have invested heavily in data centers powered by Nvidia’s advanced chips, while leading Chinese chip manufacturers like SMIC struggle with supply constraints due to U.S. sanctions on semiconductor manufacturing equipment.
Ultimately, the likely outcome is a scenario where both Beijing and Washington directly fund their domestic AI industries, each aiming to lead the race to artificial general intelligence (AGI). The stakes are immense, prompting expectations of substantial U.S. government spending to enhance technological advancements and widen the tech gap with China, which could further exacerbate the U.S. budget deficit and increase federal debt in the coming years.
Tyler Durden
Wed, 11/05/2025 – 18:50
