|
|
|
|
Science Technology and the Economic Future edited by Susan Raymond
William J Spencer
President and Chief Executive Officer
SEMATECH
Based on remarks delivered to the New York Academy of Sciences on March 10, 1996.
The last decade has witnessed significant change in the organization and conduct of industrial research and development. As a result, industry is different and the current industrial profile has important implications for the United States and its global competitiveness.
A Changed Industrial Profile
In the late 1950s, graduating from college with a degree in physics guaranteed anyone a job. Indeed, upon graduation, I received immediate job offers from companies with whom I had not even interviewed. U.S. basic research organizations were the envy of the world. Huge industrial laboratories at GE, RCA, and Bell meant equally huge basic science research opportunities, and the sheer level of effort involved created confidence that the resources would be available to pursue innovative ideas. By the mid-1970s, Bell labs alone employed some 25,000 scientists and engineers. But all of that has changed. Many of the major industrial labs are gone or much reduced, and industrial R&D is now a more dispersed corporate function. Why has the current massive downsizing of industrial R&D capacity occurred?
First, costs escalated, and companies could not afford the concentrated research power housed in many of the earlier labs. Second, markets themselves became global, and U.S. companies had to compete with overseas companies that did not have basic R&D activities, and did not have to carry that cost into their product price, nor maintain a focus on the long-term at the cost of short-term product development. If your research dollars are going into basic science and your competitor is putting all of their dollars into product development, you are at a distinct disadvantage.
Finally, technology today moves at the speed of light. The most important publishing outlet is no longer Applied Physics Letters. It is e-mail. Large corporate research entities had a difficult time being agile enough to capitalize on their own basic research with sufficient speed.
The Importance of Industrial R&D: The Case of Semiconductors
Who cares? Why does it matter that basic research in U.S. companies is declining? Anyone who bought GE stock in 19G5 when the Schenectedy Lab was changed certainly has nothing to complain about. Shareholders have done quite well. But evidence is overwhelming that industrial R&D capacity is critical to continued economic advance. The semiconductor industry offers a case in point.
The information age has come about in no insignificant part because productivity in the semiconductor industry has grown at 25% a year for the last 30 years. Every year you can buy a function for 25% less than the year before. Such productivity growth has driven not only expansion in the personal computer industry, but in communications and every other sector in which electronics plays a role.
In short, the $200 billion semiconductor industry has fueled a trillion dollar electronics industry. Currently, the semiconductor industry is growing at 15% a year, which means it will double every five years. If this trend were to continue consistently, semiconductors would be a $2 trillion industry by 2010. These growth rates may or may not be achieved, but growth at some level is sure to characterize the future of the industry. In the United States, it is important to realize that this industry was driven nearly entirely from industrial laboratories. The discoveries, innovations, and manufacturing technologies were produced by the likes of Bell Labs, Fairchild Labs, IBM, Texas Instruments, Phillips, and other companies, some of which do not even exist today.
Yet by the mid-1980s, the U.S. semiconductor industry was on its way to becoming an endangered species. The industry established several cooperative efforts in research, and pursued partnerships with universities. But progress was difficult. In 1987, the industry formed SEMATECH with federal matching funds to reestablish the nation's semiconductor dominance. Today, the industry has won back market share, and there is no federal money in SEMATECH.
So given the growth of the industry and its ability to deal with hard times and strengthen its global position, what conclusions can be drawn from the semiconductor case about the state of R&D in the United States?
First, and most importantly, we as a country do not need to spend more money on R&D. Even though Japan plans to double government-funded R&D over the next five years, and even though they will then outspend us on a dollar basis (albeit not on per capita or GDP basis), we do not need to pour more money into R&D. We should benefit from the Japanese investment wherever we can. But the $100 billion we spend on industrial research and the $70 billion in government-funded research is probably enough.
What is needed is a plan to redirect that investment. Moving funds out of defense research provides an opportunity to expand support of basic science in non-defense areas. But any attempt to redirect R&D funding will require a road map to priorities. Moreover, the opportunity is before us to develop a priorities road map for basic sciences: physics, chemistry, mathematics, and biology. It is not acceptable for scientists and engineers to argue that science can not provide such a prioritization, that the progress of science is and must be serendipity. If scientists and engineers do not determine and articulate priorities in basic science in the process of redirecting R&D investments, then senators, representatives, cabinet secretaries, and other agencies will.
A unique opportunity is before us to cooperate across disciplines and industrial groups on pre competitive science and technology that will provide innovation into the next millennium and secure a firm competitive position in the global market. The precedents are impressive. The textile and apparel industry has created and pursued such collaborative priorities; the automotive industry has; the semiconductor industry has. All are on growth paths. It is time to take this path in other areas of science and industry, to reduce the costs of technology development, and hence continue industry growth.
A Cooperative Illustration of the Case for Optimism
One final illustration makes the point. One of the factors behind the growth in productivity of the semiconductor industry is the progression toward larger wafers. A wafer is a circular disc of silicon on which are located several hundred microchips. In the late 1950s wafers were one inch in diameter. Soon they will be twelve inches in diameter. Obviously, you can make many more chips on a twelve inch wafer than a one inch wafer. In fact, today's microprocessors and memory chips would not be possible on one inch wafers. But an investment is required. One estimate is that conversion from 200 millimeter to 30 millimeter (from 8 inch to 12 inch) wafers will cost about $10 billion, not including factories or equipment. The new production factories themselves will cost about $2 billion each. No one company controls that level of investment resources. Such huge investments will take cooperation, both in terms of developing the technical capacity and in terms of financing the ultimate operation.
For the first time, such cooperation is a reality. Thirteen international companies—three from Korea, three from Europe, one from Taiwan, and six from the United States—are collaborating over the next two years to develop the equipment necessary to manufacture microchips on the larger wafers. If such collaboration is possible in an industry as competitive as the semiconductor industry, and if we can bring together cultures as diverse as Korea, Taiwan, Europe and America, R&D cooperation can be a reality in many other settings.
Email this citation