Is MCC Obsolete?

This story is from Texas Monthly's archives. We have left it as it was originally published, without updating, to maintain a clear historical record. Read more here about our archive digitization project.

The bulk of the research done by the Microelectronics and Computer Technology Corporation, better known as MCC, takes place in a pale, grayish building on the shoulder of a hill in North Austin. The structure looks down impassively at commuters traveling by on MoPac, but at the same time, it has assumed the cast of a fallen symbol: In little more than a decade, MCC has slipped from the heights of its early promise to an inglorious near demise.

When first erected, in the early eighties, the building served as a marker for patriotism, nationalism, and economic security. MCC was created by a coalition of American computer and semiconductor companies who pooled their limited research budgets so they could work together to develop new technologies. Their collective effort was supposed to stymie the thriving Japanese electronics industry, rescue its foundering American counterpart, and by virtue of its location, inaugurate another Texas boom. Yet the image did not take long to tarnish. After only a few years, MCC's funding began to dry up, as some original member companies dropped out and others cut back their participation. By the mid-eighties, when Austin's business climate went into a tailspin, local residents were grumbling about unfulfilled potential. Last summer a jury in a trade secrets and libel lawsuit awarded $26,750,000 in damages against MCC to a small chip manufacturer called Staktek and its founder—the largest award to a plaintiff in the history of Travis County. During the trial, MCC's general counsel, Tom Kirkland, protested that the judgment would bankrupt the consortium. Fast on the heels of the lawsuit came management turmoil: In a last-ditch attempt to save the organization, MCC's board began a radical reassessment of its mission, and when the dust settled, its CEO, Craig Fields, was gone.

MCC and Staktek have settled their dispute for an undisclosed amount, averting for now the threat that the consortium will close its doors. MCC has also recruited a vigorous new CEO, John McRary, who has been charming employees and shareholders alike. "Somebody once said, ‘Bad news is more interesting than good news,’ and it's always been that way," McRary says in a North Carolina drawl that should sound as slow as molasses but doesn't because of his forcefulness and peripatetic mind. "The real story at MCC is this attempt we’re going through to redefine the consortial mission in today's world. It's the story that's preoccupying me, one hundred and eighty percent of my time. I haven't found anything that you couldn't fix." His enthusiasm is contagious. "He's the best hope they’ve got," says Austin business consultant R. Miller Hicks, a sometime MCC adviser. "He has a good technical background, but he also has an ability to talk eyeball to eyeball with the chief executive officers, and he has breathed life into the employee morale situation out there."

Still, the moment when the building on the hill in North Austin might come to stand for some shining triumph seems elusive. MCC was supposed to represent a reinvigoration of the national economy, but instead it has come to symbolize the tremendous difficulties of collaborative research. In a country adamantly focused on the next quarter, any kind of long-term research and development is always a gamble. The computer industry in particular reinvents itself every six months. U.S. computer companies no longer cower before a Japanese juggernaut, but given the quickening speed with which products come and go, it has become increasingly hazardous to dedicate funds to long-term research. Whether consortia in general will work well in the new environment is an open question—one much on the minds of industry leaders. Whether McRary proves to be the savior that MCC desperately needs could determine the answer.

When MCC was conceived in 1982, Ronald Reagan was serving his first term, the Berlin Wall still stood, the Soviet Union was intact, and the United States faced seemingly insurmountable challenges from Japan. The Japanese electronics industry had decimated its competition, toppling entire companies and shaking America's sense of itself. Desperate to solve the next riddles of microelectronics before their overseas rivals could, ten U.S. computer and semiconductor companies—including Advanced Micro Devices, Motorola, and Digital Equipment Corporation—decided to combine their funds and form a research consortium, on the theory that the money would go further that way. Each member contributed a onetime fee of $500,000, which entitled it to shares in the for-profit organization, and subsequently subscribed to particular research projects at an additional cost. (The membership fee was briefly raised to $1 million but was later dropped to $250,000, where it stands today.) For an annual fee of $25,000, other companies could join as associates, which meant they could not participate in research projects but were given limited information about MCC's activities.

The main visionary behind the consortium was William C. Norris, the founder and former chairman of Control Data. "It was to better compete with the Japanese, principally, by expanding research and development, the underlying basis of innovation," Norris recalls. "The idea was that through collaboration on what we call pre-competitive research, we would have a much larger reservoir of knowledge to draw upon." Norris recruited his partners by reminding them of the Japanese plan to build a so-called fifth-generation computer, a machine that was supposed to approximate the manner of human thought. The presumed significance of this threat was needed to galvanize the companies into action, as they had a tremendous reluctance to share information among themselves. Hundreds of research consortia exist today, but in 1982 MCC was unique, and executives of the first member companies had no familiarity with this type of institution. Congress had to rewrite antitrust laws to accommodate its birth.

Back then, Texas was still basking in the effects of an oil boom, but farsighted legislators could see that if the state was to avoid serious economic hardship, it would have to acquire a more diversified industrial base. Microelectronics was viewed as an environmentally clean industry and one of the few fields that was growing instead of contracting. "The semiconductor is the building block for almost all of the computer and telecommunications revolution," says Ben Streetman, a professor of electrical engineering at the University of Texas at Austin. "Silicon is to this phase of economic development what steel was in the last century and early part of this century. It's really the foundation on which everything else rests."

The consortium's founders first thought about settling in San Antonio after a persuasive sales pitch by Henry Cisneros, who was then the city's mayor, but proximity to UT's main campus made Austin more attractive. After Cisneros alerted then-governor Mark White to MCC's interest, Texas politicians and business leaders mobilized an all-out blitz. Ross Perot lent his plane to shepherd MCC officials around the state. UT's Board of Regents promised to make available twenty acres in North Austin and team up with private contractors to construct a $22 million building there; MCC would be asked to pay only $1 a year in rent. Local banks floated subsidized mortgage loans for MCC employees, corporations kicked in the use of their jets to recruit researchers, and private businesses pledged interim office space. Austin beat out more than one hundred other cities under consideration. No city had offered nearly as much, prompting Andrew Young, the mayor of Atlanta at the time, to say, "I gather Texas bought it." Austin's power brokers were delirious over their coup. "In the year 2000 we’ll look back on this event today and we’ll speak of its economic significance in the same breath as the location of the university and the Capitol in Austin," crowed Neal Spelce, then a public relations executive, to the Austin American-Statesman.

Having spent so much time and money on getting MCC here, local residents expected to receive something in return. In this sense, the consortium was probably doomed to fail, because the hopes pinned to it by Texans hungry for economic salvation were so outsized. MCC was established to benefit the companies that founded it, as well as those that joined later, and most of those entities weren't located in Texas. By design, its accomplishments were supposed to flow out of the state—although this point was lost in the hoopla surrounding MCC's creation. "There was really no clear perception, when the recruitment was made, of what was supposed to result," says MCC spokesman Bill Stotesbery. Mark White and former Austin Chamber of Commerce president Ben Head, two members of the team that courted MCC, likewise recall that the only benefit pledged by MCC officials was a heap of good publicity.

That promise came true. Although Austin had attracted its first semiconductor manufacturer in 1974, not until MCC began its nationally publicized search for a home eight years later did the city cement its reputation as a high-tech mecca. Virtually the same team that lobbied for MCC subsequently won Sematech, a successful consortium dedicated to manufacturing better semiconductors, and many semiconductor manufacturers and their suppliers have followed, including Applied Materials, Motorola, and Advanced Micro Devices. Now some industry analysts say Texas has surpassed California as the country's largest manufacturer of semiconductors. By drawing attention to the region, MCC's decision helped generate this momentum. "I think the PR of the deal alone was worth all the money expended," says White. "Anything we got above that was gravy."

The public could be forgiven for not sharing his opinion. Most Austin residents heard an earful about MCC when it was first formed, and then very little for the next decade. The consortium's members considered its research proprietary and didn't want fanfare about its work, but many people assumed the worst—that they weren't hearing about accomplishments because there weren't any. This wasn't really true. Researchers at MCC were often productive, even if not to the degree that members had hoped. But the sense that something was amiss at MCC was not entirely mistaken.

With the benefit of hindsight, the direction of the initial research agenda entrusted to MCC became questionable almost as soon as the consortium opened its doors. The computer industry was moving to build smaller and smaller machines, not larger and larger ones, and the threat of the Japanese megacomputer proved insubstantial. "In retrospect, I think we made some mistakes in the beginning by taking such a general approach to research—almost an unfocused approach," says Roy Kuntz, a former official of NCR when it was an MCC member company. MCC has been in search of a mission ever since.

Most of MCC's research over the past ten years falls into two categories, usually referred to as hardware and software, in the field of microelectronics. It is a realm that is at once esoteric and everyday. Every time you take money from an automatic teller machine, consult a digital watch, or pop a movie into your VCR, you are using a microchip. A microchip consists of tiny transistors built onto the surface of a semiconductor. Good semiconductors exhibit a split personality: They can be coaxed into conducting electricity very well or very poorly, a characteristic which allows engineers to use the material as both a conduit for electricity and insulation against current. This convenience permits them to crowd as many as a hundred million transistors onto a single chip. One uncommonly good semiconductor is crystallized silicon, which grows in long ingots that look like oversized salamis. The ingots are sliced into wafers and cleaned in acid baths; technicians then use etching and photolithography techniques to cover them with minute transistors. Finally, the wafers are cut into individual chips with a diamond saw.

Throughout MCC's history, its hardware division has been by far the most productive division in the organization. The consortium has focused on advancements in packaging, which protects a chip against dust and other environmental hazards, and interconnect technologies, which allow electricity to flow to and from the chip. Both are almost as important as chip design itself in making electronic devices smaller and faster. (To use a crude analogy, one way of making parking lots more efficient is to build smaller cars, but without enough entrances and exits, you’ll have a traffic jam.) From 1984 to 1989 the hardware division was supervised by Barry Whalen, who, unlike many other MCC researchers, had industry experience—he was employed for twenty years at TRW, a multibillion-dollar high-tech firm, where he worked on satellite control systems and intercontinental ballistic missiles. Whalen's group improved the manufacturing and design of multichip modules (a number of chips packaged together), created a more compact wiring board on which to affix the multichip modules, and helped develop a process known as tape-automated bonding, which uses devices that look like 35mm movie reels to attach chips to a sophisticated type of flexible plastic.

The software side of the consortium, by contrast, has been far more troubled. There was, for example, the LISP debacle. A major goal of the software division was to create tools for mapping advanced integrated circuits. In the past an engineer would use a roll of butcher paper and colored pencils to map circuits, but since the task can now involve more than a million transistors, computerization helps. From 1985 to 1988 LISP (List Processing Language) was used in 70 percent of MCC's computer-aided design programs because when the consortium was formed, LISP was considered the language of the future. "It could accelerate your ability to develop quick prototypes," says Jerry Werner, who was formerly in charge of technology transfer for the computer-aided design program. "Statements in LISP could substitute for multiple statements in another language." Unfortunately, none of MCC's members ever adopted LISP, and the language failed to catch on elsewhere in the computer industry. Before MCC's researchers could rescue their project, a brand-new industry to produce software design tools rendered their efforts redundant. "Computer-aided design was the closest thing we had to a failure," concedes MCC spokesman Bill Stotesbery.

To be fair, MCC's software projects have not always been so off the mark. The consortium has developed intelligent software that prowls the Internet, the worldwide computer network, and brings back information for a cyberspace directory. It is also helping the State of Texas set up a network to link its various agencies. And many corporate clients are using MCC's neural net software, which is especially good at such thorny tasks as deciphering handwritten characters.

Aside from obvious successes and failures, MCC has occasionally indulged in truly high-risk research of the sort that commercial companies almost never support. The knowledge base Cyc, for example, may be an utter waste of time, or it could change the world. At the moment nobody can say which; still, MCC has funded the project for ten years. When Cyc's director, Doug Lenat, was recruited from Stanford University in 1984, he decided that if he articulated enough of the assumptions that human beings take for granted and programmed them into a computer, the computer could exhibit something like common sense. A computer with common sense, he reasoned, could probably understand directions in ordinary spoken languages such as English—something computers are not good at, because of the mushiness of human syntax. "I got a bunch of people who were very good at introspection and clear thinking to express the millions of things that make up common sense," Lenat says. "We did a lot of different things. We kept diaries about all the things that you do in a day. We took field trips to stores and looked through the Yellow Pages to get the names of everyday objects. We interviewed small children. Finally, we gave up on most things except introspection." Lenat's team began writing a program that would tell a computer what he calls "all the grubby little things we know about the world." He asserts that in about a year, Cyc will be able to "read" on-line versions of children's books and encyclopedias.

By and large, however, the research MCC produced did not live up to its members’ expectations. "We spent in the early years more than sixteen million dollars and didn't get anything for it," says Leonard Weisberg, the vice president for manufacturing and engineering at Honeywell, a founding member of the consortium that no longer participates in its projects. "The place was very badly managed. I would go to a meeting and ask what they expected to accomplish, and no one had an answer. They had wonderful ideas about how they were going to revolutionize the world, but they couldn't tell you what they would have by the end of the year. They hit a couple of singles, and maybe even a double, but no triples or home runs."

Because the story of MCC is about the strained relationship between the consortium and its members, there are two sides to the matter. On occasion, MCC fumbled its research, but nearly as often, member companies failed to ensure the research would be of use or to use good research once it was accomplished. In a perfect world it would be axiomatic that good research would be put to good use, but many times researchers at MCC worked diligently only to discover that their work—quite possibly of sterling quality—died stillborn.

Throughout MCC's history, its greatest difficulty has been bridging the gulf between its employees and its members. Even within a single company, a huge divide traditionally separates those who do basic research from those who translate technological advances into marketable products. At a consortium, these natural hostilities are exacerbated because the researchers are often strangers, employed at a separate organization hundreds of miles away. "I learned pretty early that I was having trouble getting through to the companies about what we were doing," says Bob Inman, the former CIA director who served as MCC's first CEO. "We had meetings of the scientific advisory panel, and we had a lot of written reports flowing out of the program, documenting what we had done, but people weren't reading them."

Inman is a genial, almost impish man, full of a boyish enthusiasm for high-tech gadgetry. During his reign, MCC acquired a reputation as a researcher's paradise where projects had distant horizons and there was little pressure to produce something immediately marketable. Longtime employees recall Inman's tenure with a special fondness, but the seeds for later problems were sown under his watch. MCC's founders had anticipated that getting technology to travel from the consortium to the members would be their central challenge. They had originally intended for technology to return to member companies with the people who invented it: Members were supposed to loan employees to the consortium for several years; when the employees returned home, they would bring their ideas back and explain them to their colleagues. But it didn't work out this way, at least not on a large scale, because the members couldn't be persuaded to part with their best researchers. So Inman chose to hire from elsewhere.

"If you’re at the mercy of somebody sending you who they can spare, you’re not likely to get the talent," he says. "The researchers were the ones who could be spared, not the ones that would do first-class research. I kept stalling, rejecting. Finally, we went outside for ninety percent of the researchers who were hired." Inman tapped smart people, but some of the new employees came from academia and were unfamiliar with the culture of the industry. "We were at arm's length from the shareholders. We didn't have a close relationship with them," says former researcher Jerry Werner. "To a great extent we were isolated or insulated from the end-users of the product. First of all, a lot of times good researchers like to pursue research that interests them. But also, the companies were all competitors. It was difficult for them to share very detailed problems with us, out of fear that their competitors would get wind of it."

MCC would send off reports about its progress, and the reports would sit on a shelf. If the reports were read, it was often by a relatively low-level employee with little access to the top brass. Bringing a product from research prototype to market requires that the idea be sold to many people, up a long chain of command, and rarely did anybody at the member companies put that much clout behind an idea that originated at MCC. "A company would join for millions of bucks," says David Gibson, the co-author of a recent book about MCC called R&D Collaboration on Trial, "and a guy would read a research report and something would actually click, and he would say, ‘I should go down and talk to those people.’ But he couldn't get the travel money to leave his company—even from Dallas, in one case, to come down [to Austin] for two days."

Many attempts were made to resolve the impasse—committees were formed, meetings were held—but they rarely succeeded. "We were naive," says Laszlo Belady, who oversaw the software division. "Let's face it. These were bureaucracies. They would say, ‘Let's send somebody down to review what those people in Texas are doing.’ Whom do you send? It should be a development engineer, yet sometimes it's another researcher. But the researcher is not a friend of MCC's; he is a competitor. He thinks, ‘Why are we spending the money on MCC? Why doesn't it go to me?’ " There were happy exceptions; one member company, NCR, regularly sent visiting scientists to the consortium in search of technology it could use. But NCR's approach was not the rule.

In 1986 Inman resigned after getting MCC up and running, as he had planned to do, and Grant Dove replaced him. A 28-year veteran of Texas Instruments, Dove sympathized with the member companies, who were restive and demanding results. In part, the newly aggressive attitude flowed from developments that were external to MCC. Radical shifts in the computer industry were shaking the consortium's founders, who had invested heavily in the mainframe business only to see it fall apart. They were forced to downsize, and as Gibson says, "It was probably easier to cut back on MCC than to lay off Joe down the hallway."

MCC's congenial, relaxed atmosphere began to change. Dove shortened the timetable on many research projects, reorganizing the work so it would be more responsive to the members’ needs. In a way, this made things less grand than they had been under Inman—the research was less "pure"—albeit more realistic. But Dove had little interest in and little funding for new projects. He had to spend most of his time cajoling old members to remain aboard and scrambling for new members to bolster MCC's precarious financial situation. In 1988, to recruit new associates, the consortium began to allow non-members to participate in research projects along with the members. (More recently, in 1993, it reduced the annual fee for associates to $5,000 for small businesses and $2,500 for universities.) In addition, MCC sought funding from government contracts. In the early years MCC was supported entirely by private industry, but today almost 40 percent of its budget comes from government sources. With the exception of fees paid by government agencies that are MCC associates (such as the National Security Agency), all government funding is won through a competitive process.

Dove introduced one other change that was to prove momentous: MCC began generating spin-offs. Spin-offs evolved when researchers took the fruit of their study—which had been paid for by the members—and formed independent start-up companies to turn their ideas into actual products. Different researchers requested permission to form spin-offs after spending years on solid work only to find that the members didn't want to provide the additional investment necessary to commercialize the technology. Dove was reluctant to let the researchers go, fearing a talent drain, but he made exceptions when the circumstances warranted doing so.

Katherine Hammer was the first out the door. Multinational corporations need to move large quantities of data from one data base to another, and that process can be slow and clumsy, so Hammer created software that performs the translation automatically. She did the basic research while at MCC, but none of the companies that sponsored her work wanted to bring it to market, and she grew frustrated. "Who wants to see shelf-ware?" says Hammer, who is charming and polished and has the slightly harried look of a woman doing it all. "If I’m going to do the research, then I want somebody to derive benefit from it. That's just my pragmatic soul." In 1990 she became president of Evolutionary Technologies Incorporated in suburban Austin. ETI made $3.5 million in its first full year of sales and twice that amount in its second.

Next to go was Steve Redfield, a gentle, dreamy physicist who was part of an MCC team that tried to build a very fast data base machine but discovered that its speed was always limited by the sluggishness of the magnetic disk drive. To get around the problem, some researchers built "Bubba," a massive computer with one thousand disk drives; others, including Redfield, tried to replace the drive entirely. Redfield looked at earlier work on holographic storage and devised a means of using a laser to store information in a transparent crystal. To develop the process, Redfield founded Tamarack Storage Devices, which was funded in its first year by MCC. After that, MCC wasn't interested, so Redfield split off from the consortium. Holographic storage has recently begun attracting a lot of attention because it can be faster or cheaper or hold more information than magnetic disk drives.

By mid-1990, around the time that two other spin-offs, Corporate Memory Systems and Pavilion, were being formed, Dove made plans to leave MCC, eventually departing in 1992. His replacement, Craig Fields, was a champion of spin-offs. This position put him on a collision course with MCC's member companies. The consortium was formed with the idea that it would send breakthroughs back to the members, not that it would spawn entrepreneurial progeny. Not surprisingly, the members soon began to feel neglected.

On paper, Fields seemed like a good fit for MCC. His previous job had been at the Defense Department's Advanced Research Projects Agency, where he made a name for himself as an advocate of government sponsorship of private-sector research. Yet almost immediately upon arriving, he rubbed people the wrong way. Aside from his bedside manner, or lack of one, he ruffled feathers by terminating some long-standing research projects, and his attachment to spin-offs quickly began to cause tension.

Fields viewed spin-offs as a potential source of revenue because of licensing arrangements that guaranteed payments to MCC in the event of profits. He liked the fact that spin-offs kept MCC's research from lying fallow and that their entrepreneurial bent gave the consortium's researchers a better sense of the rigors of the marketplace. "I reached the conclusion—which I discussed with the board, and they ended up agreeing unanimously—that we needed to make it possible for our members to get access to the technology at a point of maturity that was greater than we had been giving them in the past," says Fields. "Getting greater maturity requires greater investment. So we set up a system that basically said that our members would have the first chance to put in that extra investment. If they didn't want to, then they accepted the notion that outsiders could. Hence we got to spin-offs."

Some say Fields became so enthusiastic about the idea that he got carried away. "Craig sort of made it a mass movement," says software specialist Laszlo Belady. "He definitely overdid it." Fields also began to show an interest in subjects far afield of MCC's habitual province. Mark Eaton, formerly in charge of the consortium's International Liaison Office, which tracks technological developments overseas, recalls that Fields once casually inquired about the possibility of buying real estate in Moscow.

"If you’re asking did anyone have mixed feelings about [spin-offs], I’ll put myself at the top of the list," Fields says. "You bet there are pros and cons. The main pro was giving a commercialized source of advanced technology to the members, without additional investment, and providing some financial incentive to those of the staff who chose to engage in it. On the con side, there was always a concern that it would be so much of a distraction for people doing it, myself included, that there would be too little emphasis on doing the fundamental R and D mission of the enterprise." To mitigate that risk, Fields tied employee compensation to customer satisfaction ratings, so that if the fundamental research was neglected, researchers’ bonuses would go down.

In September 1992, with the unanimous approval of the board, Fields formed MCC Ventures, a subsidiary that was supposed to license and commercialize research conducted at MCC and also by outside parties. The first outside company recruited by MCC Ventures (the process became known as a spin-in) was the small chip-manufacturing firm RTB Technology. RTB's co-founder, Emory Garth, claimed to have pioneered a way to put more chips into a single package by stacking them on top of each other. MCC and RTB signed a joint enterprise deal in September 1992. Ten months later, however, Garth's former partner, Carmen Burns, the head of the small chip manufacturer Staktek, filed suit against RTB, and two months after that he filed suit against MCC too. Burns alleged that Garth was using proprietary information developed during their collaboration in manufacturing RTB's chips—in effect, that he was pirating Burns's research. A jury agreed. Although nobody employed by MCC was accused of direct involvement in this misuse of intellectual property, the consortium was found liable for RTB's transgressions because of the joint enterprise. Shortly after Fields left last March, the jury voted to award Burns more than $26 million in damages.

Even before the verdict in the Staktek suit, other problems began to vie for the attention of MCC's directors. The organization's general counsel, Tom Kirkland, reportedly advised the board that Fields had voted to award himself stock options in one of the spin-off companies. An outside law firm was called in to investigate. The findings have not been made public, but according to Terry Davis, an attorney who represents Joseph Koelzer, the former chief financial officer of Tamarack Storage Devices, Fields and other members of Tamarack's board voted to award themselves the option to purchase 250,000 shares each, although the deal unraveled before it was ever ratified because MCC began its investigation. (Koelzer is now suing MCC and Fields and Tamarack for wrongful dismissal, alleging that Fields orchestrated his firing from Tamarack because he participated in the investigation.) Fields is cagey on the subject. "I own stock in no spin-off," he says. "I own stock in one company that MCC was affiliated with; it has been fully ratified by the board. That's all there is to it." What about in the past? "I’ve been offered stock in other things, but I’m really not going to get into it. I think this is silly stuff."

In his defense, Fields made sure that lots of people at MCC owned stock in spin-offs. He handed out equity like candy because he thought financial incentives would help give MCC's researchers a better appreciation of market forces. Still, if Fields had attempted to own equity in a spin-off, he could have caused MCC's member companies to wonder if he was still able to keep their interests foremost in his mind.

By last spring, it was clear to many board members that the consortium's focus on entrepreneurial ventures was creating all sorts of headaches. "There was a feeling that MCC was moving into an area that it wasn't really prepared to move into," says author David Gibson. "So there was sort of a big question mark: ‘Where is it all going to go?’ When the legal suits hit, I said, ‘Well, that's where.’ " Fields acknowledges the link between venture activity and litigation. "There is a definite trade-off," he says. "By getting into these more product-oriented activities, you get benefits and you get the extra liability. What you should appreciate is that companies in the product business are embroiled in lawsuits every day. That's sort of normal stuff. Great companies like Motorola are being sued all the time. Whether MCC should be in that position is up to the board."

Aside from the danger of litigation, the emphasis on entrepreneurial activity started to threaten the definition of MCC as a consortium. Everybody who worked there suddenly seemed consumed with making his own fortune. "The more Fields pushed MCC Ventures, the more the researchers assumed that their work would go into the spin-off pipeline," says Stotesbery, MCC's spokesman. "It began to become a divisive structure, one that fostered a difference of perspective between the researchers and the members. MCC Ventures began to affect the direction of some of our research projects."

And so, over the course of the past two years, MCC's board called a series of special meetings. Larry Hollatz, who represents Advanced Micro Devices on the board, says the directors were concerned that MCC's revenues had continued to dwindle—down to about $40 million a year from a high of $70 million. They were also disturbed by the myriad problems caused by the new emphasis on spin-offs. "My concern was that we were losing good people to Ventures," says Russ McNaughton, who represents 3M on the board, "and some of the management was more concerned with Ventures than they were with the main projects. There was kind of a ground swell among the members, who felt, ‘Hey, we’re getting off track.’ " Ultimately, the board decided that in attempting to correct MCC, Fields had changed it too much.

None of the individuals present at the board meetings will speak publicly about what happened next. According to the MCC grapevine, the directors discussed whether Fields should step down "voluntarily" or be ousted. MCC's current management says only that Fields inspired a lot of bad feelings and that the rumors surrounding his departure are ill founded. At any rate, Fields was not fired because he preempted matters: He resigned last March, stating he wanted to spend more time with his family. He declines to elaborate, saying, "We put out a press release. It said exactly why I resigned. I’m not going to add anything."

MCC's new director, John McRary, had barely settled into his new office, where the windows open onto a view of MoPac and the surrounding high-tech sprawl, when the bad news about Staktek was delivered. "It was a real surprise, the way it came up on me," he says. "I knew the suit existed, but it was after I had accepted the job that this preliminary ruling came out. A lot of people called me up and said, ‘Gee, is this going to affect you?’ No. When I came down to talk to the employees, I said, ‘These legal things have a way of working themselves out, and the best thing we can do is focus on things we can do something about.’ "

McRary has iron-gray hair, a long face, and a tendency to furrow his brow. He often drops the first words of his sentences to get right to the heart of the matter. He is as restless as a two-year-old: He throws his long frame into an armchair, draws his left leg in, jogs his knee up and down, and drapes his arms along the back of the chair. The next instant, his arms are waving all over the place. He leans forward, stays there a moment, and falls back into the chair again.

McRary earned a Ph.D. in physics from North Carolina State University at age 24, and even that wasn't soon enough. "I thought I was going to go straight out of school and onto Social Security," he says. In fact he went straight to Cape Canaveral, in Florida, where he fell in love with defense work. "I worked at the Cape during the really good years, when things were exciting, when they were building all the major weapons systems for the country—Minuteman, Polaris, Poseidon. I liked the fact that you were working on things that were relevant. That you were working on problems where the solution made a difference."

For 24 years, McRary worked in the Washington, D.C., office of Science Applications International Corporation, an employee-owned firm that does research and development, primarily for the federal government and often on military systems. Most recently, he served as its executive vice president and the vice chairman of the board of directors. Bob Inman also sits on that board, and after Fields announced his departure from MCC, Inman recommended McRary for the job. McRary was interested.

"With the Clinton administration's emphasis on industrial and government partnerships, I saw the whole business paradigm of this country shifting," McRary says. "As you moved from being defense-driven more toward commercially driven, a lot of R and D work was going to shift to partnerships. There are going to have to be partnerships between government and industry, state and local, universities. So I was really intrigued with the idea of coming down here."

One of the first things he did upon arriving was to dissolve MCC Ventures, though he does not rule out the possibility of forming spin-offs in the future. He acknowledges that his central challenge will be to make sure MCC is doing research that member companies will want to use and that the research gets transferred to them. "My impression walking in, new kid on the block, is that probably MCC has not paid enough attention over the years to the tech-transfer and commercialization efforts. That's something I’m going to fix. And there are a lot of ways to do that. One of the simplest ways is to use the model that was originally intended, to have lots of assignees from the member companies. Sematech uses that model quite well. That's probably the primo way to accomplish tech transfer."

McRary plans to use studies to determine exactly which subjects the consortium should research. "I’m convinced that when you jump into a research area, you need to have built your collaborative base before you begin to spend big bucks," he says. "You need to have thought through exactly what you’re going to try to research and how you’re going to do the commercialization. We had a number of global companies who thought we needed to have a program in natural language and speech recognition. Now when I look around at all the work that has gone on in this country over the years—hundreds of millions of dollars—in that area, I say, ‘Hey, what difference would a few million bucks at MCC make?’ This is an excellent area for us to do a study in. There's an obvious demand, and yet, just jumping in and spending money is probably not a sensible thing to do."

Listening to McRary talk, it is hard to remember that he has arrived at a particularly dark moment in MCC's history. The internal turmoil, litigation, and resulting bad press make funding even harder to find and good researchers harder to hire. MCC has proved valuable to Texas by serving as a catalyst for growth and to the country by inspiring other companies to come together, and skeptics think the consortium should have rested on these laurels—that it should have declared victory and closed down a long time ago. But MCC's supporters believe that to quit now would be the worst tragedy of all, because the painful mistakes of the last decade would have been for nothing. At least, the argument goes, MCC should have the chance to learn from its missteps.

The question that has yet to be answered is whether that chance is worth taking. Are MCC's problems peculiar to its unwieldy structure? Are they intractable and woven into the competitive way we do business? Or is it possible, as McRary believes, that new management can set things right? For now, the only thing that seems certain is that this is MCC's hour of reckoning.