In early November, the Chinese government made a game-changing decision in the global technology industry.
From now on, all new data centres in China, whether funded by the state budget or using public resources, must run exclusively on domestic AI chips.
American accelerators such as Nvidia’s A100 and H100, as well as models from AMD and Intel, are effectively excluded from any project the state considers strategically important.
This is a systemic intervention affecting all government platforms based on artificial intelligence – from health and security systems to research networks and supercomputers for data processing.
The aim is not merely to replace the chips but to redirect the country’s technological trajectory at a time when American companies have become inaccessible due to the export embargo.
This sector forms the core of China’s digital statehood. Data centres serving the government projects are not simply places to store data; they are the backbone of all major AI systems Beijing is developing for public functions, from traffic identification and digital health to large-scale models for analytics, scientific research, and technological development.
In this context, a chip is no longer just a technical component – it becomes an instrument of control over the infrastructure that shapes the functioning of the state.
Therefore, this decision should not be viewed as economic protectionism for the domestic industry but as a strategic move to keep key technology fully under Chinese jurisdiction.
A response to years of US restrictions on advanced chips
At first glance, the move appears to be radical protectionism. However, it is actually a response to years of US restrictions on the export of advanced chips to China.
Since 2022, the US has blocked the export of Nvidia A100 and H100 accelerators, which are crucial for training large models, citing national security concerns.
Later, the US also banned the export of customised versions of the chips for the Chinese market – the A800 and H800 models, which Nvidia specifically designed to comply with regulations.
This has left China without access to the chips that have become the standard in the global AI industry.
Performance has become secondary to sovereignty
For a long time, it seemed that China would not be able to respond, as domestic manufacturers lag far behind their American counterparts in terms of performance.
However, under the pressure of sanctions and with state investment, alternatives have emerged.
Huawei, despite being under US sanctions, has developed chips from the Ascend series – a range of domestic AI chips produced by Huawei, designed as a Chinese alternative to Nvidia's accelerators, and used to train and run large AI models.
These chips are now recommended as the basic platform for domestic AI projects.
They cannot replace US chips in all scenarios, but they can perform enough critical functions to keep the system running.
In other words, performance has become secondary to sovereignty.
Technological autonomy as political sovereignty
The new regulation does not mean that every foreign chip is banned in China.
Private companies can still use Nvidia hardware, provided they do not participate in government-funded projects.
However, these projects have the largest budgets and greatest strategic importance.
Any upgrades or new data centres that work with the government or public sector will have to forgo US hardware
Any upgrades or new data centres in Beijing, Shanghai, Guangzhou, and other provincial centres that work with the government or public sector will have to forgo US hardware.
If a facility is already under construction and the ordered servers contain banned chips, they will have to adapt – meaning cancelling orders or replacing equipment.
Such a directive effectively destroys the market for foreign manufacturers in the most profitable segment of China's digital economy.
From Beijing's perspective, this is presented as a security issue. If a decisive part of state data, public services, and scientific archives is processed on chips from a country that has openly identified China as a strategic threat, this is seen as a vulnerability that must be eliminated.
In this way, the domestic audience is increasingly convinced that technological autonomy is equivalent to political sovereignty.
Technology becoming a strategic resource
The consequences of this decision extend beyond China's borders. Technology, which until recently was an invisible infrastructure, has now become a strategic resource around which political disputes are fought and sanctions imposed.
China has already removed American equipment from its 5G networks and is gradually replacing foreign software platforms with domestic solutions.
China is now advancing to the most critical stage - the hardware that drives AI systems.
This is rapidly forming a separate technological space in which Chinese chips, standards, and software are developed independently of American and European systems.
China is constructing a technological framework that increasingly diverges from that of the West - Huawei HQ
There is a business logic behind this. For years, Nvidia generated between a quarter and a third of its revenue in China.
If this decision is consistently implemented, that market will be lost as the competition that China is now fostering begins to win real contracts and resources.
This is the case with Huawei's platform, which already powers private and public AI services, including speech models, medical algorithms, system tools for industry, and commercial cloud portals.
At the same time, China is constructing a technological framework that increasingly diverges from that of the West.
The standards it develops in chips, software, and AI platforms are becoming less compatible with American and European ones.
Although domestic chips remain weaker than those produced by Nvidia and AMD, the political decision is to use them in strategic sectors and even to replace more technically advanced products.
This results in the creation of two separate technology systems: one based on American components and global standards, and the other on Chinese equipment and state regulations.
This division is not only about the choice of chips but also about who sets the rules and how the technological future is shaped.
Strategic instruments of power
It has become clear that determining who manufactures and controls chips is not merely a technical matter but a strategic choice with long-term political consequences.
A country that controls the key infrastructure for artificial intelligence development gains advantages in economics, security, and influence.
For this reason, chips are now regarded as strategic instruments of power, comparable to energy or military technologies, rather than simply industrial goods.
China is deliberately accepting the risk of slower technological progress to develop its own chip industry and reduce dependence on American companies
China is deliberately accepting the risk of slower technological progress to develop its own chip industry and reduce dependence on American companies.
This decision is not driven by isolationism but is a strategic choice to keep key technology under domestic control.
At a time when artificial intelligence is considered a matter of national security, the authorities in Beijing believe that technological autonomy is more important than short-term efficiency.
The conflict over chips is no longer fought in the marketplace but at the level of state strategies.
China has now made it clear that it does not wish to play by the existing rules, but to set its own.
The West can still rely on technological superiority but it can no longer rely on Chinese dependence. This is not merely an economic issue – it is a change in the way power is built and distributed in the digital age.
