The global surge in artificial intelligence is redirecting semiconductor orders to Chinese foundries as international competitors focus on producing high-margin AI chips and high-bandwidth memory, leading to a shortage in mature-node semiconductors. This shift is creating a demand crunch for power-management and other mature chip capacities, according to Zhao Haijun, co-CEO of a leading Chinese contract chipmaker. Consumer electronics and IoT companies are increasingly turning to mainland China for capacity, driven by the rise of electric vehicles, robotics, domestic supply-chain localization, and stockpiling efforts. Market data indicates that the utilization rate for older generation 8-inch wafers at the world’s top foundries is projected to reach nearly 90% by 2026, up from 80% in 2025, due to the growing need for power management integrated circuits in AI servers. As a major Taiwanese semiconductor manufacturer plans to reduce its mature-node capacity, Chinese suppliers and second-tier foundries are stepping in to meet the demand. The Chinese chipmaker’s utilization rate rose to 93.1% in the first quarter, with revenue from China accounting for nearly 89% of its total.
Is the capacity crunch in mature node chips pushing orders to Chinese foundries?
