Abstract

"Moore's Law as a 'Belief System': Long-term Dynamics of Global Semiconductor Industry (1950s-2020s) "

Takafumi Kurosawa, Kyoto University (kurosawa.takafumi.6e@kyoto-u.ac.jp)

Although semiconductors have received much attention in recent years due to their strategic importance, no geographically and temporally comprehensive industry history has been written after the numerous works of the 1980s.
This paper, which constitutes one chapter of The Oxford Handbook of Industry Dynamics, identifies the characteristics that define the dynamics of this industry (industry heterogeneity), as well as why and how industry boundaries and structures within the industry change and how this has altered the competitive factors in this industry.
Adopting an industrial history perspective, this paper will cover the global market and divide the 75-year history of the industry into three periods, identifying the content and drivers of change from the following five perspectives: 1) Technology, 2) Users and markets, 3) Intra-industry structure, 4) Geography and players, 3) Geopolitics, politics and policy. This paper is intended as a theory informed analytical overview and draws on existing research in related fields, published materials from private- and public organizations, and newspaper articles etc.
The analysis reveals the following: The most fundamental factors that defined the dynamics of the semiconductor industry were the versatility of semiconductor’s functions and the outstanding uniqueness of the product, namely continuous improvement in product performance (Moore's Law), and the resulting specialization and segmentation of the industry. Initially, a high percentage of "captive" production was done by semiconductor users, but soon specialized companies (integrated device manufacturers) became dominant in the U.S. (this was not the case in Japan and Europe), and supporting industries were derived. In the 1990s, however, the division of labor between foundries (Taiwan) and fabless firms (U.S.) was established, and IDM-centered Japanese firms lost their competitiveness. 2010s saw further differentiation by value chain and product type, and global geographic concentration.