million over 5 years, to solve common manufacturing problems and to regain U.S. competitiveness in the semiconductor industry that had been lost to Japanese industry in the mid-1980s12. In the committee’s view, SEMATECH played a strong role in early efforts to reclaim U.S. semiconductor manufacturing leadership and has been a successful example of a U.S. consortium demonstrating the value of federal funds and federal participation. This position is reiterated by a 2002 National Research Council report, Government-Industry Partnerships for the Development of New Technologies, which found that the SEMATECH partnership directly contributed to the global competitiveness of U.S. industry, specifically the resurgence of the U.S. semiconductor industry.13
Today, the SRC also continues to play a significant role in advancing the semiconductor industry though synergetic industry and university research programs and support initiatives around the world, such as the Global Research Collaboration Program, Nanoelectronics Research Initiative, Focus Center Research Program, and Semiconductor Research Corporation Education Alliance. The National Nanofabrication Infrastructure Network (NNIN) also provides a successful example of U.S. government (National Science Foundation) support of university semiconductor research. By paying for some expensive semiconductor research equipment at universities, the NNIN enables leading-edge research, which indirectly supports the U.S. semiconductor industry with research results and science and engineering graduates.
In contrast, two other industry-only consortia, started near the same time and for similar reasons, both failed. The Microelectronics and Computer Technology Consortium (MCC) was formed in the early 1980s as a response to Japan’s Fifth Generation Computer Systems (FGCS) project.14 Entirely funded by corporate partners, MCC worked on a wide range of technology and software projects, with early sponsorship particularly from mainframe computer companies. By 2000 the Board of Directors had decided to dissolve the organization. Another industry-only consortium, U.S. Memories was organized in 1989 to manufacture memories based on technology from IBM, to avoid dependence on Japanese vendors. However, by early 1990 the consortium members had proven unwilling to make the necessary investments, and major memory users like Apple, HP, and Sun did not participate, so the project was canceled. Thus far, consortia that include IT competitors but that do not have government leadership have fared poorly, due to a combination of mutual suspicion, lack of focus, and no real sense of urgency.
In summary, U.S. federal support and investment has historically relied upon a perception of military threat, economic decline, industry crisis, and/or loss of competitive position; and in the United States, electronic and computer consortia without both federal R&D support and federal direction have not generally succeeded. Thus, centralist technology policies that may work in nations and cultures that accept such direction readily are a poor match to the U.S. free-market model. Further, innovation policy has to reflect each country’s unique economic institutions, industry structure, and growth model.
3.1.2 Global Semiconductor Competition
While it could be proposed that some U.S. computer vendors “failed to innovate,” or “gave up the fight” to foreign competition, it is important to recognize the paired advantages and shortcomings of a free-market industrial economy, and the capacity it provides for innovation, not only in technology, but in the creation (and destruction) of whole economic sectors. U.S. capital market investors are often quick to spot and to capitalize on transformative shifts in a business paradigm, and, consequently, to move their investments in a way that often accelerates the change. Capital markets tend to value short-term quarterly profits and tend to reward or punish a company accordingly, which manifests in changes in its stock price. This has advantages and disadvantages. On the one hand, it discourages waste and encourages competitiveness. On the other hand, a short-term focus often discourages long-term thinking and R&D investment, particularly during difficult economic times. Federal R&D programs and public-private consortia play a crucial role in coping with this tension.
In the late 1980s, when it appeared that focused government programs in Japan, as well as unfair or unreasonable trade practices, might overtake U.S. competitiveness, DARPA investments, especially SEMATECH, drove the necessary R&D efforts in process and equipment to sustain Moore’s Law and to maintain the confidence of capital markets. Concurrently, IBM began to accelerate its investment in very high-performance semiconductor technology and to form joint innovation partnerships with numerous (non-
12See www.sematech.org/corporate history; www.sematech.org/corporate/timeline; NRC, 2003, Securing the Future: Support to the Semiconductor Industry, Washington, D.C.: The National Academies Press (available online at http://www.nap.edu/catalog.php?record_id=10677#toc).
13NRC, 2002, Government-Industry Partnerships for the Development of New Technologies, Washington, D.C.: The National Academies Press (available online at http://www.nap.edu/catalog.php?record_id=10584).
14Kazuhiro Fuchi, 1984, Revisiting Original Philosophy of Fifth Generation Computer Systems Project, FGCS, pp. 1-2.