Access Official Technical PDF

Share Summary:LinkedInEmail

FinFET technology is a notable example of an engineering breakthrough shaped not only by transistor design, but also by the relationships among universities, industry and government. Developed through research at the University of California, Berkeley, in the late 1990s, the FinFET offered a three-dimensional approach to extending the established metal-oxide-semiconductor field-effect transistor (MOSFET). Its history shows why technical advances can be difficult to understand without looking beyond the design itself to the networks that support research, development and adoption.

Why the semiconductor industry needed new transistor designs

For decades, semiconductor development relied on making MOS transistors smaller so that more could fit on an integrated circuit. This scaling supported increasing computing capacity in a limited physical area. But reducing transistor dimensions indefinitely was not feasible. As gate lengths became shorter, it grew increasingly difficult for the gate to control the flow of electricity through the channel.

The industry had substantial reasons to continue improving MOS technology. Manufacturers had invested in production equipment, processes and workforce expertise built around it. Moving to a fundamentally different transistor type could require extensive changes to that established infrastructure. The challenge was therefore not simply to find a technically promising design. A successor also had to be compatible enough with existing industry knowledge and manufacturing capabilities to attract support.

How FinFET technology extended the MOSFET

A conventional planar MOSFET controls a channel in a two-dimensional layout. FinFET technology adds a vertical dimension: the channel takes the form of a fin, and the gate controls it from multiple sides. This arrangement provides a different way to manage the transistor channel as planar devices approach scaling limits.

The underlying idea had earlier precedents. Research on three-dimensional MOS designs, including work published by Digh Hisamoto while at Hitachi, and earlier investigations into trench transistors contributed to the wider intellectual context. But an initial concept is not the same as a working device. The Berkeley team’s work involved developing and fabricating test transistors, drawing on theoretical research as well as practical expertise in areas such as lithography.

The project grew out of a 1997 call from the Defense Advanced Research Projects Agency (DARPA) for research on devices below 25 nanometers. The Berkeley proposal, led by Chenming Hu with colleagues Tsu-Jae King and Jeffrey Bokor, included a three-dimensional version of the MOSFET. Research and development continued from 1996 to 1998, with contributions from graduate students, postdoctoral researchers, visiting industry personnel and staff, alongside the principal investigators.

Innovation networks connected research to industry

The FinFET was developed within a broader shift toward cooperation in semiconductor research. In the 1980s and 1990s, U.S. companies faced increasing international competition and the rising costs of research and development. Industry groups, research consortia and government programs helped coordinate longer-term work that could be difficult for individual firms or universities to support alone.

These arrangements connected industrial priorities with academic research and public funding. University labs could pursue longer-term questions, while industry provided guidance, expertise and links to potential users. Nearby facilities, including Stanford and Lawrence Berkeley National Laboratory, provided additional research resources. Researchers also moved between academic and industrial settings, helping information and practical knowledge travel across institutional boundaries.

That brokerage role mattered. Patent co-inventorship networks mapped by the study show the Berkeley researchers connecting clusters of inventors associated with different companies and universities. In contrast, comparison networks around IBM project leaders were more concentrated within IBM. The maps do not capture every relationship or prove how knowledge moved, but they help make otherwise difficult-to-see connections visible.

Why continuity helped the design gain attention

FinFET’s three-dimensional structure presented manufacturing challenges, but it remained a MOS transistor technology. That continuity mattered in an industry with significant investment in MOS-based processes and equipment. The design proposed a new way to extend a familiar technology rather than requiring an immediate move to a completely different semiconductor platform.

Close contact between researchers and industry also helped keep the work connected to industrial concerns. The Berkeley team presented its work to companies, and the technology attracted industry interest. Intel announced in 2011 that it would move toward using three-dimensional transistors. The interval between the research and that announcement illustrates that an engineering concept and an industry-ready technology are distinct stages, particularly when adoption involves manufacturing change and investment.

What patent and network analysis can—and cannot—show

Digital tools can help researchers examine patterns across patent records that would be laborious to trace by hand. Co-inventorship maps can show links among named inventors, organizations and periods. Patent citations can help identify earlier inventions referenced by later patents, while comparisons of patent language can offer another way to explore relationships among technical documents.

These methods have limits. A patent citation does not necessarily represent a direct flow of knowledge: citations may be added by patent examiners, and patenting practices do not capture all research or collaboration. Co-invention records also show only a particular kind of relationship. As a result, network diagrams are best treated as evidence to investigate, not as complete explanations. Oral histories, archival research and technical analysis remain important for interpreting what the connections mean.

Conclusion: a system-level view of engineering progress

The history of FinFET technology demonstrates that a major design advance can depend on more than an inventive idea. Research funding, university facilities, industrial input, skilled contributors and established manufacturing pathways all helped shape how the technology developed and gained attention. For technical decision-makers, the broader implication is that evaluating an innovation requires attention to both its engineering characteristics and the ecosystem needed to develop, validate and adopt it. Combining qualitative evidence with patent and network analysis can provide a more complete, while still appropriately cautious, view of that process.

Access Official Technical PDF

About The Author