What is a 'domestic chip' in the context of China's AI data centers?

A 'domestic chip' refers to an integrated circuit designed, and ideally manufactured, within China. This includes chips developed by Chinese companies like Huawei or Alibaba, which are designed to perform the complex calculations needed for artificial intelligence. The goal is to reduce reliance on foreign designs and fabrication facilities, ensuring a supply chain that is not vulnerable to external geopolitical pressures.

How will China fund this $295 billion AI data center initiative?

The funding for China's ambitious AI data center buildout is expected to come from a mix of sources. These include sovereign debt, such as ultra-long-term special government bonds, direct state funds allocated for strategic industries, bank loans, and investments from private capital. This multi-pronged approach reflects a national commitment to financing this critical infrastructure.

What role do China Mobile and China Telecom play in this plan?

China Mobile and China Telecom, two of China's largest state-owned telecommunications operators, are expected to play a central role in the new AI data center network. They are slated to take primary responsibility for running these large-scale computing facilities and maintaining the connectivity between them. Their involvement leverages existing national infrastructure and operational expertise for a rapid and coordinated rollout.

Image: courtesy of Thenextweb

techJuly 21, 2026By Veridact EditorialUpdated Jul 21

China's AI Ambition: Building Data Centers Without Nvidia

A Chinese artificial intelligence lab, Z.AI, has completed a large data center powered entirely by Chinese-made chips, a significant step that signals Beijing's efforts to circumvent US export controls are yielding results. This development arrives as China outlines a massive $295 billion, five-year plan to construct a nationwide network of AI data centers, with an explicit goal of using 80% domestic chips. The initiative aims to bolster China's AI industry and reduce its reliance on foreign technology, intensifying the global competition for AI supremacy.

Outlook

The completion of Z.AI's data center, operating without any Nvidia hardware, marks a tangible achievement in China's long-term strategy for technological self-sufficiency. This is not an isolated event but part of a much larger, nationally coordinated effort. The Chinese government, led by senior agencies like the National Development and Reform Commission, is working to integrate fragmented regional computing facilities into a unified national network. This infrastructure is intended to be the backbone of China's AI development, treated as a strategic national resource akin to electricity grids or transportation networks.

Funding for this ambitious $295 billion buildout is expected to draw heavily from sovereign debt, including ultra-long-term special government bonds, supplemented by state funds for strategic industries, bank loans, and private capital. State-owned telecommunications giants, China Mobile and China Telecom, are slated to take primary responsibility for operating these facilities and ensuring their connectivity. This centralized approach aims to replicate the success seen in past campaigns that fostered national champions like Huawei, now with the objective of replacing US technology across the entire AI stack.

The immediate consequence is a clearer path for domestic chip manufacturers, such as Huawei, and AI companies like Alibaba, which has already launched a data center in China utilizing its own AI chips. This concerted push indicates that while US export controls have created significant hurdles, they have also accelerated China's resolve and investment in indigenous solutions. The next five years will likely see a rapid expansion of these domestic-chip-powered data centers, presenting both opportunities and challenges for Chinese tech firms as they scale their production and innovation capabilities.

Background

The global race for artificial intelligence has increasingly become a contest over computing power, with access to advanced AI chips sitting at its core. For years, Nvidia has dominated this market, supplying the high-performance graphics processing units (GPUs) essential for training complex AI models. However, escalating geopolitical tensions, particularly between the United States and China, have led to stringent US export controls on advanced semiconductor technology. These controls aim to limit China's access to cutting-edge chips and prevent their use in military applications, effectively creating a technological blockade.

In response, Beijing has made technological self-reliance a national priority, pouring vast resources into developing its own semiconductor industry and AI ecosystem. This push is driven by a recognition that dependence on foreign technology creates significant vulnerabilities, especially in critical areas like AI, which is seen as fundamental to future economic growth and national security. The current initiative to build AI data centers with domestic chips is a direct manifestation of this strategic imperative, designed to mitigate the impact of US restrictions and foster a homegrown, vertically integrated AI supply chain. The plan is not merely about replacing foreign hardware but about building an entire ecosystem of software, hardware, and infrastructure that is resilient to external pressures.

Precedents

China's current strategy for AI self-sufficiency echoes previous national campaigns designed to cultivate domestic industrial champions and reduce reliance on foreign technology. A prime example is the rise of Huawei Technologies Co. Decades ago, Huawei, initially a relatively small player, benefited from significant state support, preferential policies, and a protected domestic market, allowing it to grow into a global telecommunications equipment giant. This model involved strategic investment, directed research and development, and a focus on building robust domestic supply chains, even if it meant slower initial progress.

Similarly, in other strategic sectors, the Chinese government has historically used a combination of state-backed funding, policy directives, and market protection to nurture local industries. This approach often involves setting ambitious targets, allocating substantial capital, and encouraging collaboration between state-owned enterprises, research institutions, and private companies. The explicit goal of achieving 80% domestic chip usage in the new AI data centers, coupled with the $295 billion investment and the involvement of state-owned entities like China Mobile and China Telecom, aligns directly with this established pattern. It indicates a long-term, top-down commitment to building an indigenous AI 'stack' – from the foundational chips to the large-scale computing infrastructure – rather than a short-term reaction to trade restrictions.

The shift by China to build its AI infrastructure using exclusively domestic chips carries profound consequences for the global technology industry, national economies, and the future of artificial intelligence development. For US chipmakers, particularly Nvidia, it signals a potential long-term loss of market share in one of the world's largest and fastest-growing AI markets. Even if Chinese chips are not yet on par with the most advanced US offerings, the sheer scale of China's planned $295 billion investment ensures that domestic alternatives will rapidly improve and find widespread adoption within China.

This move also represents a significant step towards a bifurcated global technology ecosystem. As China pushes for self-reliance, it will inevitably create a distinct set of hardware and software standards, potentially leading to two parallel AI development paths. This could complicate international collaboration, fragment research efforts, and force global companies to choose sides or develop separate offerings for different markets. For consumers and businesses, this could mean variations in AI product performance, compatibility, and availability depending on their geographic location.

Crucially, by treating computing power as a strategic national resource, China is signaling its intent to control a fundamental element of future economic and military power. This strategic independence in AI infrastructure could grant China a significant advantage in areas like data privacy, national security, and the development of AI applications tailored to its specific policy objectives. The real stakes here are not just about who makes the fastest chips, but who controls the foundational technology that will drive the next generation of innovation and global influence.

Scenarios

Analysis

[{"title":"Accelerated Chinese AI Independence","description":"China's concentrated investment and strategic coordination could significantly accelerate its progress towards AI independence. The $295 billion commitment, coupled with the 80% domestic chip target, creates immense demand and incentive for local chip designers and manufacturers. This could lead to rapid advancements in Chinese-made AI chips, potentially narrowing the performance gap with international leaders over the five-year plan. Success in this area would solidify China's position as a self-sufficient AI superpower, lessening the impact of future export controls and enabling it to pursue its AI development trajectory unhindered by external technological dependencies. This outcome would force global tech companies to adapt to a Chinese market increasingly dominated by domestic hardware and software, potentially leading to a more fragmented global tech landscape."},{"title":"Increased Global Supply Chain Fragmentation","description":"The determined push by China to build its own AI infrastructure is likely to further fragment global technology supply chains. As China prioritizes domestic suppliers, demand for non-Chinese components within its borders could diminish, impacting revenue for international chipmakers and hardware providers. This could lead to a 'decoupling' effect, where distinct supply chains emerge for different geopolitical blocs, increasing costs and complexities for companies operating globally. Furthermore, the development of unique Chinese AI chip architectures and software stacks could create interoperability challenges, potentially leading to different standards and ecosystems for AI development worldwide. This fragmentation could also spur other nations or blocs to similarly invest in domestic capabilities for critical technologies, further balkanizing the tech sector."},{"title":"Challenges in Scaling and Performance","description":"While China has demonstrated the ability to build data centers with domestic chips, scaling this effort to a $295 billion national network with high-performance requirements presents significant challenges. Domestic chip performance may not immediately match the efficiency and power of leading international alternatives, potentially leading to higher operational costs or slower AI model training times. Production bottlenecks, yield rates, and the complexity of designing and manufacturing advanced semiconductors at scale could also slow down the ambitious rollout. Furthermore, the integration of diverse domestic hardware and software components into a cohesive, high-performance national network will require sophisticated engineering and could encounter unforeseen technical hurdles. This outcome would mean China achieves its self-sufficiency goals, but perhaps at the expense of cutting-edge performance or economic efficiency in the near to medium term, leaving a persistent gap with the most advanced global AI capabilities."}]

Timeline

2017-12-13

Google Announces China AI Lab

Google announced plans to open an artificial intelligence research lab in Beijing, signaling early interest in China's AI talent and market, predating the current heightened geopolitical tech tensions.

2026-07-20

Z.AI Unveils Nvidia-Free Data Center

A Chinese AI lab, Z.AI, completed a large data center running entirely on Chinese-made chips, without any Nvidia components. This is confirmed as a clear sign of Beijing's progress in bypassing US export controls.

2026-07-20

Alibaba Launches Domestic Chip Data Center

Alibaba CEO Eddie Wu announced the creation of a technology committee he will head, and the company launched a data center in China powered by its own domestically developed AI chips, aligning with national self-sufficiency goals.

Ongoing (as of 2026-07-20)

China Plans $295 Billion AI Data Center Buildout

China is drafting a five-year plan to spend $295 billion on building a national network of AI data centers. The plan aims for 80% domestic chip usage and is being led by senior agencies like the National Development and Reform Commission, with China Mobile and China Telecom taking on primary operational roles.

Frequently Asked Questions

US export controls are government regulations that restrict the sale of certain advanced technologies, particularly high-performance semiconductors and chip manufacturing equipment, to China. These controls are designed to prevent China from acquiring cutting-edge chips that could be used for military modernization or to gain a technological advantage. For China's AI industry, these controls mean a limited or no supply of the most advanced AI chips from companies like Nvidia, forcing Chinese firms to develop their own alternatives.

Discussion

Be the first to share your thoughts.