China and the United States may be competing for dominance in artificial intelligence, but some Chinese manufacturers are finding opportunity in the very infrastructure race that is intensifying the rivalry.
For companies that make the hardware and modular systems behind data centers, the U.S. AI boom is not just a geopolitical flashpoint. It is a commercial opening. "We can see very big potential in the U.S. market," said S.K. Lee, a global vice president with Singapore-registered Brightray, an AI infrastructure company. "The China market is equally important, but we can see that there's a stronger demand in the U.S."
Brightray's sole manufacturer is Chinese firm PrefabDC, which produces prefabricated data centers, while Brightray manages construction of the buildouts. The arrangement reflects a broader reality that is often obscured by the political rhetoric surrounding AI: China remains deeply embedded in the physical supply chains that make the digital economy function, even as the two countries compete for technological supremacy.
The scale of the U.S. buildout helps explain why Chinese suppliers are paying close attention. The United States has more AI data centers than any other country, with 5,427 in 2025, according to data published by the Stanford Institute for Human-Centered Artificial Intelligence. China had 449, according to the same report. At the same time, U.S. technology giants including Alphabet, Microsoft, Meta and Amazon are estimated to spend about $765 billion combined this year on AI infrastructure. JPMorgan Chase CEO Jamie Dimon recently said that figure could reach $1 trillion next year.
That level of spending dwarfs what Chinese companies have committed. Alibaba and other Chinese firms have announced far smaller investments, while Beijing has outlined plans to spend $295 billion on data centers over the next five years. Even so, that remains far below the pace and scale of private-sector investment in the U.S.
Jeffrey Ding, an AI expert and assistant professor at George Washington University, said the gap is rooted not only in capital but in market demand. "Chinese companies just are not generating as much revenues from their AI services," he said. "And that's one of the reasons why they're not able to invest in these enormous data center buildouts that their competitors in the U.S. are doing." Ding added that the deeper issue is that "there is not as much demand for the AI services."
The U.S. market's appetite for AI infrastructure is creating a pull effect that reaches well beyond American borders. Chinese companies are already supplying parts of the data center ecosystem, including transformers, batteries and fiber-optic cables. Benjamin Boucher, principal analyst on supply chains for Wood Mackenzie, said the U.S. does not have enough domestic capacity in key areas and faces long lead times for some components.
"China is definitely very important towards the U.S. supply chain for data centers at the moment," Boucher said. He described the environment as increasingly difficult because of tariffs, restrictions and the broader deterioration in bilateral ties. "It is a very challenging environment between the geopolitics recently between the tariffs we've been seeing and the recent bans," he said. "I think it is just going to depend on what we see at a political level over the coming years in terms of how it evolves."
That political uncertainty is central to the business case. American policymakers and security officials worry that reliance on Chinese suppliers for critical infrastructure could create cybersecurity vulnerabilities and strategic leverage for Beijing, especially as the U.S. seeks to maintain an edge in AI. The Trump administration is considering bans on Chinese open-weight AI models as well as new restrictions on Chinese data center components.
Yet the supply-chain logic remains powerful. The U.S. is facing shortages in key electrical equipment, and domestic production cannot always match the speed or scale required by hyperscalers racing to expand capacity. China, as the world's largest manufacturer, can often deliver faster and at lower cost. Boucher said China can offer those components "at a much more favorable timing."
At a factory in Yangzhou, an eastern Chinese trading city, Lee described how Brightray benefits from that industrial base. Materials and equipment are sourced from multiple parts of the world, but a large share comes from China's dense manufacturing network, which has spent decades building the logistics, labor and supplier ecosystems needed to support large-scale industrial production.
That advantage is now colliding with a more hostile geopolitical climate. Xi Jinping's visit to Washington has renewed attention on the trade relationship, and AI featured prominently in talks between Xi and President Donald Trump. But the broader relationship remains strained, with tariffs, technology controls and national security concerns shaping the outlook for companies on both sides.
For Chinese manufacturers, the U.S. AI data center boom presents both promise and peril: a massive market hungry for equipment, and a political environment that could tighten at any moment. For now, the demand is strong enough to keep them looking west.