Abstract:    The current severe international competition pattern has made China′s semiconductor industry face unprecedented challenges, and improving its technological innovation efficiency has become the focus of discussion by experts and scholars from all walks of life. Based on the relevant data of technological innovation input and output of 76 listed enterprises in China′s semiconductor industry from 2019 to 2022, this paper classified the listed companies in the semiconductor industry into five links: design, fabricate, assembly and test, equipment and materials, and used the BCC-DEA and Malmquist-DEA models to conduct a static and dynamic evaluation analysis and investment redundancy analysis of the technological innovation efficiency of the whole industry and all links of the industrial chain. The results showed that: (1) From the static perspective, the overall technological innovation efficiency of listed enterprises in China′s semiconductor industry is not high. Although pure technical efficiency is improving, the number of DEA effective enterprises is declining, less than 1/5 of the total. The average value of technological innovation efficiency in design, equipment and materials is higher than the industry average, and the technological innovation efficiency in fabricate and assembly and test is relatively weak. (2) From the perspective of dynamics, the total factor productivity of China′s semiconductor industry fluctuates greatly and the overall technological innovation efficiency shows a downward trend, while that of fabrication and assembly and test shows an overall growth trend, and that of design, equipment and materials show a downward trend. (3) From the perspective of investment redundancy, the redundancy of R&D personnel and fixed assets investment is more prominent, and that of R&D expenditure is not significant. This paper will supplement the quantitative research on technological innovation in the semiconductor industry, further expand its research perspective, and provide practical empirical basis and policy recommendations for optimizing resource allocation in the semiconductor industry and improving the efficiency of technological innovation.

Key words: semiconductor industry, technological innovation efficiency, BCC-DEA model, Malmquist-DEA model, evaluation