skip to main content

The Last Word

Sematech: A Look Back

David Lammers

When Sematech was folded into SUNY Polytechnic in mid-May, it brought to mind how challenging Sematech’s existence has been, and how a single person—in Sematech’s case, industry-legend Bob Noyce, co-inventor of the integrated circuit—can make such an outsized difference in pulling people together behind a common goal.

Robert (Bob) Noyce

In the early 1980s, Japan had stunned the U.S. semiconductor industry with its quick rise in memories, having mastered two essential technologies: CMOS-based DRAMs, and lithographic equipment (steppers). Noyce went there on a personal fact-finding mission, where he observed the high quality of Japan’s semiconductor manufacturing fabs. He came back home and warned Gordon Moore and others that the United States had to improve its manufacturing or face further reversals.

Noyce also wrote a personal check for $50,000 to Larry Sumney, who was trying to get the Semiconductor Research Consortium (SRC) on its feet. He urged Sumney to use the money to pay rent and other expenses while Sumney lobbied the U.S. industry and Washington for funding.

Sematech was formed in 1988, and a Japanese assessment then gave Sematech only a 25% chance of succeeding, given American cultural propensities. Were it not for Noyce, that prognosis would have proved correct. Almost immediately, a problem arose. The 14 member companies were obligated to send assignees to Austin; however, many of those assignees were not the best talent available. One of the senior managers at Sematech in those early days much later confided that he had been instructed to make sure that none of his company’s secrets leaked out.

With Sematech stumbling right out of the gate, Noyce offered to step in, becoming the organization’s CEO in mid-1989. Immediately, young engineers from member companies eagerly sought out assignment to Sematech and a chance to rub shoulders with Noyce.

The U.S. Congress had agreed to put up $100 million each year for 5 years, matching the financial contribution of the member companies. To make sure its money was being well spent, a full-time staffer from the Government Accounting Office was on site in Austin to monitor Washington’s investment. And there was plenty of grumbling from members of Congress about favoring one industry over another—a shortsighted view.

Besides attracting talent and securing financial support, Noyce and the succeeding Sematech leaders faced another challenge: how to devise a research agenda that would satisfy the member companies. By 1990 most of the U.S. industry had abandoned DRAMs, and Sematech switched from running a memory process development organization and prototyping fab to a microprocessor-centric agenda. Not long after that, Micron Technology understandably withdrew, followed by LSI Logic, Motorola, and others. The comings-and-goings of Sematech member companies had begun.

Noyce once said that his main job at Sematech had little to do with technology development. His success would be judged on changing a culture in which the semiconductor companies beat up their equipment vendors on price. Sematech’s main goal, Noyce said, was to engender a cultural change in which the chipmakers saw their suppliers as partners. In that sense, Noyce and Sematech put a business value on collaboration that continues in varying degrees today.

There are lessons to be learned from Sematech. One is that relying on state funding is not a good idea for research consortia that have a much broader focus. Mike Polcari, who ably served as Sematech’s CEO during a period when the state of Texas gradually withdrew its support, recalled how a promised grant of $35 million turned overnight into a loan, which had to be paid back, with interest. Little wonder that Sematech—which had helped put Austin on the map as a semiconductor technology center—decamped to Albany, New York, lured by a $300 million investment from that state.

A second lesson is that consortia should not be seen as employment agencies. Both Texas and New York viewed Sematech as a source of local jobs, rather than as a national center to strengthen an essential technology foundation.

Add up the challenges—stable funding, attracting talent, a broad agreement on the R&D agenda, establishing trust—and Sematech’s track record of successes and disappointments begins to come into better focus. Bob Noyce was a big enough man to get many of us to see the benefits of collaboration and invest in making it work—and a rare one, indeed.

David Lammers is an Austin-based technology journalist.