A statistical model to analyze the complexity of semiconductor manufacturing

Semiconductor chips power all modern electronic devices and gadgets, such as smartphones and tablets, as well as the massive infrastructure that supports the Internet. These chips are manufactured using progressively smaller integrated circuits that measure in the billionth of a meter, or nanometer. The fabrication methods used to develop these chips entails some of the most advanced and complex technological processes known to humankind. Semiconductor manufacturing processes have evolved enormously over the years, with each evolution harboring increasing complexity and novel innovations. The academic literature confirms that the accelerating pace of technological obsolescence necessitates the continued investments in process innovations even in the face of increasing risk and uncertainty brought about by large capital outlays that are vital to overcoming the complexity of manufacturing these chips. Using multivariate statistical analysis methods and publicly available data from a leading semiconductor manufacturing firm, a model is developed to assess the increasing difficulty of semiconductor manufacturing. The findings are consistent with the hypothesis that semiconductor manufacturing difficulty is accelerating despite massive investments in process innovations. Managerial implications are discussed along with suggestions for further study, including refining the proposed model to more accurately measure the rate of change, investigation of the impact of breakthroughs in technological capabilities and environmental factors, such as the “green” movement, and in updating the dataset and the model to comprehend other leading semiconductor manufacturing firms through survey of the experts.