Austin’s proposal for the SEMATECH site included a coordinated effort among universities, companies, and state and local governments who had decided in advance that Austin would be the only city the state would offer for consideration. The University of Texas offered to buy a former Data General plant for $50 million and then allow SEMATECH to use it at no charge. The Austin city council supplemented this with $250,000 and a promise to “cut through any red tape hindering the project.” The Texas congressional delegation, under the leadership of House Speaker Jim Wright and Representative J. J. Pickle (a member of the powerful House Ways and Means Committee), actively assisted the SIA in its federal lobbying efforts to secure funding for SEMATECH. The state of California’s proposal efforts, meanwhile, were so fractured that Governor Deukmejian bypassed the state legislature, which was drawing up its own plan, to submit San Jose’s bid for SEMATECH.20
In January 1988, the site-selection committee announced that Austin, not Silicon Valley, would house SEMATECH. Within months, construction on the fab was underway. Along with many others, San Jose mayor Tom McEnery worried that the SEMATECH decision meant the state of California was “in danger of losing our leadership position with the high-technology industries.”21
BY THIS POINT, SEMATECH had successfully assembled its “black book,” consulted with the Department of Defense, lobbied Congress, chosen a location, broken ground for a fab, determined a dues structure, and assembled a membership base of 14 companies and DARPA (the Defense Advanced Research Projects Agency)—all without anyone having a clear sense of what the consortium would actually do. There was a general agreement that a worthwhile first task would be for SEMATECH to try to manufacture a huge run of chips, not to sell, but as a research project. Engineers at SEMATECH—half of them employed by the consortium, half of them assignees from member companies – would implement and optimize the production processes for these chips, learning from each other and sharing their insights along the way. When the most efficient approach had been determined for every step, SEMATECH would disseminate this information to its member companies.
But what chips would they build? In January, SEMATECH issued a proud press release announcing that IBM and AT&T had both contributed “Manufacturing Demonstration Vehicles”—soup-to-nuts guidelines to the several-hundred-step process needed to build advanced chips—to SEMATECH. The chip that AT&T offered was brand new and had not been produced in volume, so its value as a vehicle for demonstrating mass-production techniques was questionable. IBM’s contribution was an established chip already in production, but the company, concerned about intellectual property, refused to release to SEMATECH details on a specific process step and further stipulated that only IBM assignees or SEMATECH direct hires could work with the technology. Upon hearing of IBM’s proprietary restrictions, AT&T summarily imposed similar ones. Such limitations thoroughly undermined the very industry-wide sharing and cross-company interactions that SEMATECH was supposed to foster.22
Moreover, the AT&T and IBM devices were so different that SEMATECH’s decision to accept both signalled the consortium’s overall lack of direction. IBM’s chip was a DRAM, the quintessential commodity chip that had served as a technology driver since the late 1960s, and the first chips that the Japanese built more successfully than the Americans. By contrast, AT&T’s chip was an SRAM, a device built with the same process technology used to manufacture more specialized circuits, including small lots of special-application chips. Accepting both the IBM and AT&T demonstration vehicles enabled SEMATECH to postpone a decision on whether it would focus on commodity chips or more specialized devices.
The consortium’s efforts were so diffuse, in part, because it had no leader. Three men, each of whom held a full-time, high-level management job at his own company, and none of whom was empowered to decide the direction SEMATECH should take, ran the consortium as an interim management team while a search committee vetted permanent CEO candidates. Meanwhile, almost 300 people (including the assignees from member companies) worked at SEMATECH, but there was no Human Resources department. One consultant called in to work with the consortium described the situation thus: “They stacked five hundred people out there all at once in what seemed to be a two week period, and there was no structure. Literally, there was no management structure to speak of. I’d never seen anything like it before.” Nor was there any formal method to determine how money was spent, or how contracts for research and development were to be awarded.23
A search committee comprised of Sporck, Noyce, and Jerry Sanders of Advanced Micro Devices spent weeks sifting through some 200 names of potential CEO candidates. They found it nearly impossible to come up with a name acceptable to all 14 member companies and the Department of Defense, and often the names that came up—including Noyce’s—belonged to men who were not interested in the job. The choices were further limited by the requirement, adopted at SEMATECH’s founding, that none of the top officers at the consortium could work for a member company after their tenure at SEMATECH ended. This restriction was adopted in an effort to assure objectivity and reduce the risk of raiding, but it was quite a bit to ask, given that most viable candidates would likely be found in very senior management positions at member companies to which they would like to return. On the rare occasions that a promising candidate seemed interested in the SEMATECH job, his employer had offered him a raise and a promotion. “I got more damn guys promoted,” Charlie Sporck mused a decade later.24
In April, 1988—ten months after Congress approved funding for SEMATECH—DARPA, citing the lack of “specific details on [the consortium’s] operating plan,” as well as the failure to find a CEO, announced that it was delaying the release of the federal portion of the consortium’s funding. Funding was released one month later, but then a frustrated United States Senate threatened to trim SEMATECH’s expected $100 million allotment for 1989 to $45 million on the grounds that without a chief executive, SEMATECH was falling behind schedule. These moves created even more problems for the CEO search committee. “No right-minded executive would interrupt a promising and lucrative career to commit to a project that might disintegrate beneath future legislative budget debates,” opined Electronic News.25
NOYCE FOUND THE ENTIRE SITUATION appallingly embarrassing. One day in July, he called Richard Hodgson, his mentor since the earliest days at Fairchild. “I’d like your advice,” Noyce said. He was considering volunteering for the job as SEMATECH CEO. Several people had urged him to take the position in the past, but he did not want to leave California and frankly thought he was too old to run another startup effort. In a comment reminiscent of his youthful reluctance to accept the general manager’s job at Fairchild, Noyce had urged the SEMATECH search committee to “find someone better than me.” But with SEMATECH’s federal funding under threat, Noyce had begun to reconsider. He still did not want the job, he said, but “I prefer not to see all of [my] life’s work go down in flames.” He told Hodgson, “I wouldn’t go down there, but I’m embarrassed by what’s going on and it’s just—it’s ridiculous. [SEMATECH] is a good effort.” He had told Ann Bowers the same thing.26
Hodgson advised Noyce not to take the job for two reasons: “You got married and you love to fly airplanes.” There would not be much time to travel with his wife or muse on infinity if Noyce became SEMATECH’s CEO. Noyce seemed to think his mentor had a point. Hodgson hung up the phone believing he had talked Noyce out of volunteering.27
The next day, Noyce flew to Austin, where the SIA board was meeting to discuss SEMATECH. On the flight back, Noyce sat next to his friend Charlie Sporck. The two were griping about the search problems over a few Beefeaters on the rocks when Noyce suddenly said, “You know, Charlie, I think I ought to probably be the CEO there.”
Sporck was surprised but recovered immediately. “Bob,” he said, “the job is yours.”28
Sporck thinks that “the idea of bringing all of these companies together and making them play beautiful music was appealing to Bob. … He s
aw himself as good at that.” But in a private Intel interview conducted shortly after Noyce was named SEMATECH CEO, Noyce said that he did not expect the position to offer much in the way of excitement or fun. He certainly never pretended he wanted to leave his life in California, and with a net worth in excess of $100 million, there was no reason he needed to.29
At the press conference announcing his new position, Noyce said, “SEMATECH is just too important for the industry and the country to ignore the call to leadership.” For generations, Noyce men had heeded the call to service through leadership in its explicitly Christian guise. In his own career, Bob Noyce had tuned his ear to an entrepreneurial version of the same summons. To be sure, the desire to secure his own financial and professional success had driven the earliest stage of Noyce’s career, but once he had accomplished this, he had turned to young startup companies precisely because this was how he believed he could be of best service. Not one of the entrepreneurs with whom Noyce worked ever thought he had been drawn in for the money—though Noyce welcomed that, too. Instead, he had wanted to help, to mentor, to nurture, and to give back. As one of Noyce’s friends put it, “Bob spent his life walking along the edges of his Christianity.”30
Perhaps Noyce was consoling himself or moving his decision into a realm from which he had drawn such pleasure in the past when he coined a new phrase to describe the consortium. SEMATECH, he said, was a “public start-up,” an experiment in “national entrepreneurialism, with industry and government teaming up as venture partners.” But if Noyce really did consider SEMATECH to be a novel sort of startup, then his decision to lead the effort in its time of crisis fits another pattern he had established long before. In the three decades he had worked in high-technology industry, Bob Noyce had never willingly allowed one of his startups to die.31
THE NEWS THAT NOYCE would head SEMATECH was greeted with praise and relief when it was announced on July 27, 1988. Sporck publicly rejoiced, calling Noyce, “the ideal guy with the greatest stature in the industry.” Almost immediately upon receiving the news, congressional leaders assured Noyce that SEMATECH would receive the full $100 million it had expected. Even SEMATECH’s most vocal critic was impressed. “I was afraid [SEMATECH] was going to turn into a turkey farm,” said T. J. Rogers of Cypress Semiconductor, who had repeatedly predicted that the consortium would benefit only the largest firms in the industry. “But with a guy of Noyce’s stature, that won’t happen.” One editorial said simply, “When an infant business needs credibility, there’s nothing like hiring a legend.”32
Only Noyce’s closest confidants expressed anything less than enthusiasm for his decision—not because they thought he would not be good at the job, but because they thought it would not be good for him. Before hearing that Noyce would run the consortium, several of his friends had believed SEMATECH was such a hodgepodge of strange bedfellows that it was doomed to fail. One letter from a colleague Noyce had known since Fairchild begins, “Condolences and congratulations on your new job.” The entirety of Arthur Rock’s comment to Noyce—“I think taking the Sematech chairmanship was a class act”—was not an unqualified endorsement of it. Nor was Andy Grove’s note calling him a “mensch” for accepting the position. “The U.S. government, the U.S. semiconductor industry, and Sematech are damned lucky that you are a glutton for punishment,” wrote Ken Oshman, founder of ROLM. “With you at the helm, a possibility I never contemplated, all my objections [to SEMATECH] evaporate.” Gordon Moore summarized the feelings of many of Noyce’s friends when he said the SEMATECH job was “a huge sacrifice” for Noyce. “I didn’t think Bob wanted to take on a terribly difficult job, a job with a strong political content as well as a major managerial task at the time,” Moore says. “But Bob did it, he went down there.”33
Two weeks after Noyce was named CEO, Japanese semiconductor manufacturer NEC asked to join SEMATECH. The consortium refused the application with no public comment beyond the statement that it had “reaffirmed [its] intent to keep the consortium limited to U.S. companies.”34
To have admitted a Japanese firm would have undermined the patriotic fervor that at times seemed the only thing uniting the 14 member companies and the Department of Defense. In its earliest incarnation, SEMATECH was so “tightly wrapped in the United States flag” (as one person put it) that management considered reserving its best parking spaces for American-made cars. The house banner, “The American Enterprise Flag,” reproduced the famous colonial banner of a rattlesnake (but with 14 rattles) coiled above the phrase, “Don’t Tread on Me.” There had been talk of making this flag the focal point of SEMATECH’s first annual report—the cover shot would have shown men in bunny suits raising the flag Iwo Jima-style—but Noyce, who approved of the sentiment, nonetheless squelched the idea as “too over the top.”35
Noyce had originally planned to live in California and commute to Texas—he looked forward to the flying—but the sense that he needed to “lead by example” and commit himself fully to the job, coupled with the lack of a personal income tax in Texas, swung him to declare official residency in Austin. He and Bowers moved into a nice, but not extravagant, split-level home in August. After taking a previously scheduled trip to Tibet, Noyce assumed the helm of SEMATECH in September. The official opening of the SEMATECH facility two months later was a lavish affair. After a brief performance by the Austin Symphony and a fly-over by a pair of jets from a nearby Air Force base, Noyce took the stage beneath an enormous American flag. Arrayed on either side of him were members of Congress, officials of the Reagan and Texas gubernatorial administrations, the mayor of Austin, representatives of each of SEMATECH’s 14 member companies, and various leaders of the University of Texas system—69 people in all. “Throughout the history of our nation, Americans have stepped up and met challenges,” Noyce said. “SEMATECH is the American answer to a most modern challenge.”36
NOYCE EXPECTED HIS JOB at SEMATECH to resemble the work he had done at Intel. He said that he would be “Mr. Outside,” offering testimony in Washington and speaking around the country about the importance of SEMATECH and its mission. As far as internal affairs were concerned, he said, he would be “an observer and a counselor. … When I’m here I will sit in on the staff meetings and that sort of thing. Try to contribute what I can.” He added, “I hope to be able to be a sounding board … and perhaps to be able to see some things about the organization just because I’m standing a little bit farther away from the force there [in Austin].”37
Playing “Mr. Inside” to Noyce’s “Mr. Outside” would be Paul Castrucci, a 32-year veteran of IBM, who would serve as chief operating officer. Noyce and Castrucci had never met, but industry insiders pictured the arrangement as something akin to a Noyce-Grove partnership, with Noyce providing SEMATECH’s public image, esprit de corps, and long-range vision; and Carstrucci serving as the get-things-done operations man. Noyce named his own chief administrative officer, Peter Mills, and then made the unexpected announcement that the three men would serve together in an Intel-style “Office of the Chief Executive,” rather than Noyce serving as the sole chief.38
But Noyce soon found there were too many problems within the organization for him to focus exclusively on outside relations. The assignees at SEMATECH were paid by the member companies who sent them to the consortium. It took Noyce only a few weeks on the job to realize that most people at SEMATECH “have split loyalty. Is their cause [the company that pays them and to which they will return after their stint at SEMATECH], or is their cause SEMATECH? That’s an interesting one to work around. How do we get that all pulled together?” he wondered.39
Miller Bonner, head of communications for SEMATECH, recalled that the member companies did not help the consortium’s cause. Some of them saw SEMATECH as a place to dump low-performing employees. Most of them sent their legal counsel to chat with assignees before they left for Austin. “Whatever you do, don’t open your mouth. But take a lot of notes,” the attorneys would advise. “And remember, we’re suing com
panies B and C over here, so watch what you say.” Concerns about intellectual property were so great that an early SEMATECH planning meeting had attracted 32 intellectual property lawyers from potential member companies and the Department of Defense. As Noyce put it, rather resignedly, “Everyone comes here to extract something from SEMATECH and nobody comes here to contribute something to SEMATECH.” It was a classic tragedy-of-the-commons situation. If everyone did only what was best for his or her individual company, Noyce explained, “there isn’t going to be anything to extract [from].”40
These troubles kept Noyce from performing the “Mr. Outside” job he had loved at Intel, and instead forced him to play a role similar to the general manager’s function he had grown to detest at Fairchild. SEMATECH lacked much of the interdivisional tension that had hobbled Fairchild, but the wrangling among member companies produced the same effect. The lawyers’ don’t-tell counseling proved extremely effective, and Noyce found himself pulled into the very sorts of meetings he thought he would avoid at SEMATECH. He would have to go around a conference table, individually asking each engineer present how she or he thought whatever given problem at hand could best be solved. Bonner, the communications expert, claimed that Noyce’s approach was effective. “When you’ve got somebody of that kind of stature looking you in the eye and asking for your opinion, you’re not going to say, ‘I’m not going to tell you.’ In a very short time, it became very apparent that everybody was holding back the same kind of information.”41
The Man Behind the Microchip Page 44