SMIC founder Richard Chang argued that semiconductor success should not be measured solely by the ability to manufacture chips at 3nm or 2nm. He said Chinese companies could also make substantial contributions by addressing gaps in mature manufacturing processes, specialty technologies, materials and equipment.
Chang described the view that only leading-edge processes represent success as a misconception. In the interview, he argued that developing expertise in a specific market and resolving a supply bottleneck could be more valuable than competing across every segment of the industry.
Mature processes and specialist suppliers
Chang said advanced processes accounted for less than 20% of the global semiconductor market by product quantity, while more than 80% of demand came from mature processes and specialty technologies. These were his estimates, expressed in terms of product quantities rather than sales revenue.
He said companies should identify missing capabilities in the semiconductor supply chain and choose which problems to address first. His argument covered specialized components, materials and equipment as well as chip manufacturing.
Chang applied the same reasoning to regional industrial development. He said second- and third-tier Chinese cities should develop distinctive semiconductor specialties rather than copy the broad, capital-intensive wafer-manufacturing projects pursued by larger cities. He argued that differences in funding, skilled workers and supporting infrastructure made that approach more suitable.
Leading-edge manufacturing and specialty processes both feature in TSMC’s business. In its 2025 annual report, TSMC said its 2nm technology entered high-volume manufacturing in the fourth quarter of 2025. The company also said it was working with customers to develop specialty technologies using mature processes.
Companies such as Apple and NVIDIA have described production of their chips at TSMC’s Arizona operations.
Export controls provide additional context for China’s semiconductor industry. The Netherlands expanded licensing requirements for specified deep ultraviolet lithography equipment from September 7, 2024. The Dutch government said applications would be assessed individually and described the measure as an authorization requirement rather than an export ban.
On December 2, 2024, the U.S. Bureau of Industry and Security announced additional controls covering semiconductor manufacturing equipment, software tools and high-bandwidth memory. The agency said the measures were intended to restrict China’s ability to produce advanced semiconductors with military applications.
Application-specific AI hardware
Chang also argued that edge and distributed AI applications had received insufficient attention compared with cloud-based large models and large computing facilities. He said many hardware requirements for specific applications remained unmet.
He identified industrial control, automotive electronics and wearable devices as markets needing semiconductor components and solutions tailored to particular uses. In his assessment, these applications offered opportunities for companies to develop specialized products.
Chang contrasted those opportunities with cloud AI projects, which he said required sustained, intensive investment from national platforms or very large financial groups. He argued that such projects were outside the strengths of ordinary startups and favored application-specific development as a way for smaller companies to pursue markets beyond direct competition with dominant suppliers.
