China is preparing a $295 billion plan for AI data centers aimed at excluding Nvidia.

      China aims to dominate the AI race using its own hardware. A recent proposal reveals its willingness to invest heavily and the extent to which it will go to exclude American chips from the equation.

      Beijing is creating a strategy to allocate approximately 2 trillion yuan ($295 billion) over the next five years to establish a national network of AI data centers, as reported by Bloomberg, referencing sources familiar with the situation.

      This initiative, spearheaded by the influential National Development and Reform Commission, seeks to connect the country’s dispersed computing facilities into a unified, integrated grid by 2028, primarily managed by state telecommunications giants China Mobile and China Telecom.

      A key aspect of this plan is the technology that will be utilized in these data centers. The proposal stipulates that local suppliers, including Huawei, must deliver at least 80 percent of the essential technology, including AI chips, thereby effectively excluding Nvidia and AMD from the picture. This approach mirrors previous campaigns that established national champions like Huawei, now focusing on substituting US technology throughout the AI stack and bridging the gap with American research labs.

      The funding will mainly derive from sovereign debt, including ultra-long-term specialized government bonds, as well as state resources for strategic sectors, supplemented by bank loans and private investment. This initiative is part of a larger “Six Networks” program that encompasses water, power, and computing, with the integration of the power grid potentially elevating the total expenditure to over 5 trillion yuan.

      “Elevating it to a national strategy ensures policy alignment and capital mobilization,” noted Charlie Dai, a principal analyst at Forrester.

      Despite the ambition, the figures may appear smaller compared to the West. The $295 billion spans five years, while US companies, led by Meta and Microsoft, are allocating about $725 billion for AI within this year alone. The Chinese estimate does not include private expenditures by Alibaba and Tencent, and constructing Chinese data centers is comparatively more affordable.

      The key takeaway is not just the total investment but the coordination involved: a state-led effort mobilizing debt, land, power, and chips to support a unified national grid.

      This initiative reflects rising Chinese confidence in its own semiconductor capabilities. Washington has relaxed restrictions, permitting Nvidia to sell its previous-generation H200 chips to Chinese customers, but shipments have yet to commence. In May, nine domestically produced AI chips from Huawei, Alibaba, Shanghai Biren, and Moore Threads passed a domestic security review, allowing them access to sensitive sectors.

      Beijing is increasingly convinced that it can bridge this gap independently. This notion of sovereignty is gaining traction in the West as well, with Britain’s sovereign-AI initiative and Europe’s efforts to reduce dependence on US cloud services, albeit from a different perspective: while Europe is concerned about its reliance on America, China is focused on minimizing its need for it. The plan remains in its early stages, and sources have cautioned that details may evolve. Nonetheless, the trajectory is clear.

      The world’s two largest economies are each attempting to create separate AI supply chains, signaling the end of a unified global ecosystem.