AI race comes down to power and data centres and China holds a clear advantage, says veteran investor
A leading Chinese tech investor believes the global competition in artificial intelligence is increasingly being shaped not by breakthroughs in algorithms but by the scale of the infrastructure needed to power them. Allen Zhu Xiaohu, managing director at GSR Ventures and a well known figure in China’s start up ecosystem, said during a recent podcast that China is on track to surpass the United States in AI capability within the next decade. His argument focuses on the rapid construction of data centres and the strong electricity supply network emerging across the country, assets he believes are becoming central to the future of AI development.
Zhu explained that while the US continues to lead in foundational research, China has built momentum in the areas that determine whether AI systems can be expanded and deployed at scale. In his view, the next phase of competition will belong to the country that can provide the computing power and energy needed to operate increasingly complex models.
Infrastructure becomes the decisive battleground
According to Zhu, China’s ability to catch up on algorithms and AI models is far greater than America’s ability to rapidly build new data centres and power plants. He noted that China has been able to construct large scale computing hubs at remarkable speed, often supported by regional governments looking to stimulate local technology ecosystems. These facilities are essential for training large language models, running cloud services and supporting AI driven applications in industry and public services.
By contrast, Zhu said the US faces more hurdles when expanding its infrastructure. Lengthy approval processes, higher construction costs and an electricity grid already under strain make it challenging for the US to match China’s pace. As the demand for computing power grows, these constraints could become major obstacles for American AI companies.
China’s coordinated strategy for power and computing capacity
A significant part of China’s advantage, Zhu said, lies in its coordinated national and provincial planning. Massive data centres have been built in regions rich in renewable energy, including areas with abundant hydroelectric and wind resources. This approach not only boosts computing capacity but also reduces the environmental footprint associated with energy intensive AI operations.
China’s ability to build quickly and secure consistent energy supplies has allowed companies to operate high performance computing facilities more efficiently. Zhu pointed out that China’s renewable energy expansion complements its AI ambitions, giving it access to a stable and growing source of low cost electricity.
Insights from a seasoned investor in China’s tech landscape
Zhu is often referred to as a unicorn hunter because of his track record of identifying billion dollar start ups such as Didi Chuxing and RedNote. His perspective carries weight because he has been involved in China’s technology development for decades and understands the structural factors that influence long term success.
In discussing the AI race, Zhu emphasized that the world is entering a stage where infrastructure matters as much as, if not more than, breakthroughs in engineering research. Training modern AI models requires enormous amounts of data storage, power and cooling capacity, making the physical backbone of AI increasingly important. Countries that invest heavily in these foundational layers will, in his view, shape the next wave of global technology leadership.
Implications for the future of global AI competition
Zhu’s comments highlight a broader shift in how analysts view the AI rivalry between China and the United States. Instead of focusing solely on models, patents or research output, attention is shifting toward supply chains, computing resources and long term energy strategies. As AI becomes more deeply embedded in economic and military systems, the nations with superior infrastructure will have the advantage.
China’s rapid buildout of data centres and its robust electricity network could enable it to scale AI faster and more efficiently than its competitors. Although the US still leads in innovation and houses many of the world’s top AI researchers, the challenge of meeting soaring energy and computing demands may slow its progress.
Zhu’s analysis suggests that the future of AI will be determined not just in laboratories but in the construction of power plants, renewable energy grids and high density computing hubs. His message is clear: the race for AI dominance is becoming a race for infrastructure, and China is moving quickly to secure its lead.
