The biggest risk to the U.S. semiconductor industry is to say that our problems have been solved. That assessment, by SEMATECH president and CEO William Spencer, speaks volumes about the dramatic turnaround of a vital U.S. industry.
Spencer, the first speaker in the new LBL/UC Berkeley-sponsored lecture series on Science and Technology in a Competitive World, said that when SEMATECH was formed in 1987, the fortunes of the industry were in precipitous decline.
"Industry, government, and the military feared the semiconductor industry was headed the same way as the U.S. television industry," Spencer said during his Feb. 11 noontime talk. "In 1975, we had two-thirds of the world market. By 1985, Japan had surpassed us, with our market share dropping to 40 percent."
In 1987, government and industry formed the jointly-funded SEMATECH research and development consortium. Spencer said its efforts to develop quality manufacturing techniques deserve some credit for the revival of the U.S. industry, which today sells slightly more semiconductors than do the Japanese.
Spencer said it was difficult to say how the trend had been reversed or to pinpoint SEMATECH's role, but a number of factors were involved. Korean competition, which stiffened in 1989, hurt the Japanese in terms of sales and profits. Japanese interest rates used to be much lower than rates here in the U.S. But the cost of capital has evened out, and U.S. industry is no longer at a competitive disadvantage when it comes to borrowing money to pay for new equipment. Certainly, he said, SEMATECH and prior government/industry efforts to foster university-based research and development have helped.
SEMATECH is an industry-led organization with 750 employees, about 250 of which are on loan for two-to-three years from private industry. It has a $180 million annual budget, and represents about 80 percent of the U.S. semiconductor industry. The organization has played an important role in developing new manufacturing equipment, testing equipment, and improving the quality and yield of integrated circuits.
"I believe we have shown that government and industry can work together," said Spencer. "We have accomplished together what could not have been done if either government or industry had gone it alone."
Spencer, who has headed SEMATECH since 1990, earned a Ph.D. in Physics from Kansas State University and is a member of the National Academy of Engineering. He began his career at Bell Telephone Laboratories, has held positions at Sandia National Laboratories and Xerox, and was recently named to the Secretary of Energy's Advisory Board.
He said he sees research universities playing an increasingly important role in fostering the industry but was less clear about how the national laboratories fit in.
He said that up until now industry had accounted for almost all of the major breakthroughs in the field of semiconductors. With U.S. industry scaling back their research and development efforts, he said he believed university research is a likely source of future progress. He advocated the creation of perhaps eight university-based semiconductor research centers with partial funding by industry.
He also advocated a shift in priorities and funding within the national laboratory complex, and called for a reduction in funds going into nuclear weapons research. With the Cold War over, Spencer said, nuclear weapons work can be scaled back, releasing new resources to work with industry.
"There are about 700 federal labs with a $20-25 billion annual budget," he said. "You have unique facilities and a tremendous pool of human resources at these laboratories. Now, you need to demonstrate you can work with industry and do so on a time-scale that demonstrates the sense of urgency that industry operates under."