By John Rhea
WASHINGTON - The next major challenge for Sematech, the semiconductor industry research consortium based in Austin, Texas, is 193-nanometer-resolution lithography for semiconductor masks.
The organization`s president, William Spencer, envisions pushing feature size down to 50 nanometers before radically new technology will be needed.
Interviewed at a National Academy of Sciences-sponsored roundtable on global technology strategies, Spencer said other infrastructure efforts will include new exposure methods, resists, and metrology, but that Sematech will continue to stay away from product research and development "recipes" for specific components.
The organization, which now has 10 member companies, was established in 1987 under sponsorship of the Defense Advanced Research Projects Agency as part of a national effort to achieve international competitiveness. Half of its $200 million annual budget initially came from DOD.
Sematech`s charter was to focus on the front end of semiconductor processing technology in an attempt to bolster the U.S. component industry. That original purpose has been achieved, according to Spencer, who estimates that the U.S. now controls 43 percent of the worldwide semiconductor market, followed by Japan with 37 percent, and Korea and Europe with about 10 percent each.
As a result, Congress withdrew federal support three years ago and Sematech`s budget, now provided entirely by its corporate members, has fallen to about $150 million a year.
At least two-thirds of that is devoted to front-end infrastructure with the balance going into design, test, safety, and health issues, Spencer says. While engineers at some of the member companies are working on post-silicon materials, he maintains that silicon will continue to dominate for another 25 years.
The semiconductor market is evolving, however, with the U.S. staking out a lead in digital signal processors and microprocessors (an 85 percent share of the world market for the latter) and in the chemical vapor deposition process with the international competitors dominating the low-price RAM and other commodity markets.
Principal growth will continue to be in the industrial, consumer, and automotive applications, according to Spencer, thus providing the industrial base for military and aerospace users.
The domestic personal computer market is far from being saturated, he adds, and the product life cycle of PCs for users is only three or four years. This was once common for autos but has grown steadily in recent years.
