AUSTIN, Texas -- Since its inception in 1987, Sematech has been the model for an American partnership between government and industry to cultivate a technology that ultimately benefits the entire economy.
Nearly five years later, support from both sides of that partnership is flagging.
On the government side, the proposed Bush administration budget for the coming fiscal year cuts federal support for Sematech by 20 percent, from $100 million to $80 million.
On the industry side, as many as three out of the 14 founding members may leave the Austin consortium aimed at restoring U.S. supremacy in the manufacturing of the key component of Information Age economies, the microchip.
By some estimates, that could cost Sematech as much as $15 million a year more.
The financial blows are damaging, but not deadly, contends William J. Spencer, Sematech's president and chief executive officer.
The only sure departure among its members is LSI Logic Corp., a Milpitas, Calif., maker of custom chips. One other, Micron Technology Inc., said it plans to leave in midyear. A third, Harris Corp., said it also is weighing the merits of leaving.
All three are small members of Sematech. Almost all the rest are microchip giants, such as Texas Instruments Inc., International Business Machines Corp., Digital Equipment Corp., Intel Corp. and National Semiconductor Corp. They represent the bulk of the chip-making industry and, as a result, the vast majority of the market for the manufacturing processes and equipment improvements that Sematech turns out.
"From our point of view, we have been very pleased with Sematech's contributions so far," said Thomas F. Gannon, director of planning and development for Digital Equipment Corp. He contends DEC gets access to the equivalent of $10 worth of research and development for each $1 it puts in. "We pay a large amount of attention to the leverage factor," he said.
"I don't anticipate erosion of major members," said Wally Rhines, executive vice president of TI's Semiconductor Group. "If 85 percent [of the industry] votes for what you're doing, it isn't so bad."
The Sematech chief also prefers to look at the $80 million in recommended government support as a glass that is four-fifths full, rather than one-fifth empty. Indeed, he argues that there was a strong chance just a few months ago that the glass would be totally empty.
After five years, he feared, the Bush administration might recommend no additional funding for Sematech, in light of ballooning budget deficits and declining global tensions. Sematech's federal support comes from the Defense Department.
"I would look at this as a great success," said Mr. Spencer, of keeping most of its funding in the proposed budget. "The question was, would the government support Sematech at all?"
Nonetheless, the immediate prospect for Sematech is scrimping the projects it has chosen to focus on for improving processes and equipment. Even before these impending cutbacks, its annual budget has never risen significantly above $200 million, even as inflation takes its toll.
There are few opportunities for raising new money. On the
government side, Mr. Spencer is hopeful that Texas lawmakers in Washington, such as Sen. Lloyd Bentsen and Representative J.J. "Jake" Pickle, can rally support in Congress for the full $100 million in the final fiscal 1993 budget.
But on the industry side, Sematech has not announced a single new member since its founding five years ago. Mr. Spencer also appears reluctant to ask existing members to pony up more money than they already contribute.
The reluctance is understandable, said Mr. Gannon. "I think we all recognize the financial climate that we're all going through, including the federal government. Times are tough."
Instead, Mr. Spencer said he will try to run Sematech more efficiently, to make up for what may be a 15 percent drop in its overall budget.
Since coming on board in 1990 after the death of the charismatic Robert N. Noyce, the low-key Spencer has moved to speed up progress on existing projects at Sematech. Where only 50 percent of its projects were meeting their schedule, now 80 percent are on time. That will save money.
"I have never seen a project that overran its schedule that didn't overrun its budget," said Mr. Spencer.
Mr. Spencer also said Sematech will cut low priority projects to save more money.
That's all Sematech can do, said G. Dan Hutcheson, president of VLSI Research Inc., a San Jose, Calif., firm that tracks the microchip manufacturing industry. Sematech came about because the United States is in a "catch-up mode," trying to regain ground lost to Japan.
The United States is catching up, after Sematech's first five years of cooperative efforts, Mr. Hutcheson said. One benchmark: The number of usable chips that result from the creation of tiny circuits on circular slabs of silicon, called wafers.
In the five years before Sematech was created, the United
States went from leading Japan by 10 percent in that "yield" to trailing by 15 percent, a huge swing. In the past five years, the Japanese lead on yield has dropped to 9 percent.
Overall, Sematech's greatest contribution to its members so far may be to "define a methodology" for companies to evaluate equipment and determine its overall cost of ownership. Mr. Spencer also pledges to add a new methodology this year that will allow members -- or potential new members -- to calculate clearly their financial returns on investing in Sematech.