While the US remains the leader in advanced semiconductor chips and AI infrastructure, China’s strengths in energy supply and AI applications signal a potential shift in global AI leadership over the next 10 to 20 years, according to a report by Li Cheng of the University of Hong Kong.
- US leads in cutting-edge chips and AI infrastructure
- China excels in energy capacity and AI application deployment
- China narrows AI model gap with cost-efficient, open-source approach
What happened
Li Cheng, founding director of the Centre on Contemporary China and the World, co-authored a paper assessing the AI rivalry between the US and China using Nvidia CEO Jensen Huang’s five-layer cake framework. The study benchmarks energy, semiconductor chips, AI infrastructure, AI models, and AI applications.
The paper finds the US currently dominates in advanced chips and AI infrastructure, but China leads in energy capacity and has surpassed the US in AI application integration by leveraging cost-effective, open-source AI models. The gap in AI models is shrinking as China ramps up research and development investments.
Why it matters
The US’s current AI dominance benefits from advanced technology and infrastructure but faces long-term risks from structural challenges, including rising competition in chip technology and the energy-intensive nature of AI development. China’s expanding energy supply bolsters its capacity to power large data centers, a critical factor in AI scalability.
China’s adoption of open-source AI models reduces integration costs, accelerating AI adoption across diverse sectors. This strategy contrasts with the US preference for premium, closed-source models, potentially providing China with a strategic advantage in achieving sustainable AI growth and technological leadership over time.
What to watch next
Attention should focus on China’s progress in overcoming semiconductor bottlenecks, especially in high-end GPUs and design automation tools currently constrained by US export controls. The success of China’s increased R&D investments and state-backed subsidies in these key technology areas will be crucial.
Simultaneously, monitoring how the US responds to infrastructure and energy competition challenges will be important. The escalating energy demands of AI data centers may influence both countries’ ability to maintain or expand their AI capabilities and global influence in the sector over the coming decades.