Nation’s High-Tech Think Tank : MCC Under Pressure to Turn R&D; Into Products

A massive new building hulks in the forest at the north end of this central Texas city, its top three stories jutting out like a furrowed brow over the entrance. An as-yet unfilled trough sidles up to the clay-brown structure, more suggestive of a moat than a pond.

Inside, beyond the locked doors that flank the security desk, an atrium reaches to the ceiling, its blank white walls dividing the 200,000-square-foot building lengthwise. Along either side of it, columns of computer-filled offices march down inner halls, progressively numbered but otherwise mute to the identity of their occupants or purpose.

This is the home of the Microelectronics & Computer Technology, or MCC as it calls itself in a cryptogram to the outside world. A $23.5-million fortress in the battle for technological superiority.

Bobby Ray Inman, the “superspy” who three years ago crafted the cloak of secrecy under which 21 technology companies--from Control Data to Lockheed--could pool their resources in the name of cooperative research and in the process equip America for the fight, is about to leave MCC. And while it is certain that the doors of MCC won’t soon be flung wide to spill out technology secrets, Inman’s departure signals a new, more open phase at MCC.

In recent months, MCC has begun to squirm under the glare of the outside world--from the budget-conscious companies that support it, from the community that held such high hopes for what MCC would bring to town with it and--though still an experiment itself--from the dozens of other newly formed research ventures that view MCC as a model.

Even as it responds to these inquiring glances with small doses of information about its early progress, MCC is facing its own internal pressures.

Although it has quieted in-house squabbles that threatened it during its formative months, received government sanction for its structure and gotten its research projects well under way, MCC still faces formidable hurdles: among them, handling new leadership, growth and a constantly modulating set of demands from its member companies.

But ultimately, MCC’s value will not be measured against the expectations of the outside world, or even by the quality of its research. Rather, its ultimate success will depend on whether its member companies put to use the tools and technologies that MCC develops.

“Our technology needs to be used. An army does not conquer the world without deployment,” said Palle Smidt, senior vice president at MCC.

Not the least among those who worry about whether MCC’s shareholder companies can do that is Inman, a retired four-star admiral who even 10 years ago was preaching the gospel of technology leadership while he headed the National Security Agency and then was deputy director of the Central Intelligence Agency.

Inman, who announced in September that he’ll be leaving MCC when his contract expires at the end of the year, said that when he joined MCC, his goals “were to get it clearly established, created and the research under way.” Now is a good time to leave, he added, because “there are no problems on the table among the companies.”

His choice of a new job is illustrative of his eagerness to get on with the business of technology. Once he leaves MCC, he’ll be head of a new Texas-based venture, called Westmark, that will acquire technology companies. Inman believes that Westmark will give him the opportunity to do first-hand, and in shorter time, what MCC’s member companies need to: turn technology into products.

21 Different Approaches

As for how MCC’s members are facing that challenge, Inman said: “There are 21 shareholders, and I see 21 different approaches to technology transfer. In a few companies, they’ve put in place the mechanisms to pull the technology in. They’re closely tracking it and put a suction on that pipeline. But with a very large number of others, if they are taking those actions, it’s not apparent to me.”

MCC’s board of directors, comprised of a representative from each of its 21 member companies, said that in searching for Inman’s successor, they will be looking for someone with a strong technical background--which Inman does not have--who can direct the transfer process.

But MCC can only do so much, say its executives and program directors, citing scientific equivalents of the old maxim “you can lead a horse to water but you can’t make him drink.”

The consortium has implemented several methods to accommodate the flow of the research into the member companies, including written reports, briefing sessions and liaison arrangements. If applicable, prototypes could be built and demonstrated.

There are few existing models of technology transfer in this country, including universities and government-sponsored research.

But either of those would be unworkable at MCC, because of its particular form as a consortium of competing companies, or especially in the case of government-sponsored research, of little encouragement.

“NASA, in its early years was fairly successful” in spinning off technology to the private sector, said Sheridan Tatsuno, a analyst who tracks the Japanese semiconductor industry at market researcher Dataquest and who has followed MCC’s progress. But, he said, NASA has “no systematic approach” to technology transfer and has been unable to spin off a large percentage of its work.

The issues of technology transfer are cropping up much sooner than Inman and its originators expected.

Had to Adjust Goals

Billed from the beginning as a long-term research project with a 10-year deadline for delivering results, MCC has had to adjust its goals as it added members.

The research consortium grew out of a long-held dream of William C. Norris, the founder and then-chief executive of Control Data. In 1982, it found its impetus in the U.S. technology industry’s concern over Japan’s massive, government-sponsored “fifth-generation computer” research projects.

“I worked on setting up MCC long before the Japanese appeared on the horizon as a threat,” Norris said in a recent interview. “But what made (the consortium) possible was the Japanese threat.”

Nine other companies joined with Control Data, agreeing that no one American company--aside perhaps from industry giant International Business Machines--could afford to match that effort.

A plan outlining four areas of long-term research was hatched among the 10 original participants. The four basic areas are software technology, semiconductor packaging and interconnect, computer-aided design and advanced computer architecture. The architecture program has been further subdivided into four programs: artificial intelligence, database, user-machine interaction and parallel processing (supercomputing). It is the advanced computer program which most closely matches Japan’s fifth-generation computer research.

Once Inman was selected to head the consortium, he engineered a national search for a site.

“I deliberately made the decision to do a public site selection,” he said, “partly to nail down the support of all the companies. But also because it was a unique kind of operation and it needed as much life support as we could engender, and that the community (that was chosen) saw it as potentially important economically.”

Membership Increased

By the time Austin was selected in May, 1983, over such locations as San Diego and Atlanta, the number of member companies was up to 12 and read like a who’s who of technology: Advanced Micro Devices, Allied, Control Data, Digital Equipment, Harris, Honeywell, Motorola, Mostek, National Semiconductor, NCR, Sperry and RCA.

Nine more companies have since joined: BMC Industries, Bell Communications Research, Boeing, Eastman Kodak, Gould, Lockheed, 3M, Martin Marietta and Rockwell.

The addition of new member shareholders beefed up the budget, which next year will approach $75 million. But also, new members added to the conflicting expectations and put pressures on that 10-year schedule for delivery of the technology.

“The goals of the original 10 companies were not for short-term results; everything was long term,” Inman said. “But what has happened over time . . . is there are more short-term expectations.” He explained that the high-level executives who first organized MCC were “looking at mountains,” but MCC is now being directed by people at the member companies who are looking at “what’s going to make their profit centers profitable.”

MCC decided to allow the release of its newly developed technology to the companies as soon as it could be documented and meet quality assurance standards. “I was persuaded,” Inman said, “that we’d run the big risk of the technology being irrelevant” if MCC waited the entire 10 years to let it out the doors to the member companies.

Even with that changed approach, MCC officials insist that the consortium is ahead of schedule, especially in comparison with other basic research, such as at universities. They’ll discuss such advances in general terms only, however, to protect whatever competitive edge they may gain.

Low-Key Operation

“American companies are more closed-mouth than the Japanese,” Dataquest’s Tatsuno said. “They don’t go out and flaunt their successes. (At MCC) they’ll be low key. They don’t want to attract a lot of attention and tell everybody what is hot.”

Other aspects of the original plan were unworkable in implementation and led to some of the in-fighting. It was planned for the consortium to get about 65% of its researchers from the member companies. Insiders complained, though, that the companies were unwilling to spare their most qualified researchers to MCC. Today, the percentages are flipped, with about 65% of the staff coming from the outside.

“The reality was that the talent was not in the companies, or they didn’t make it available, and I think it was the former,” Inman said.

In the staffing of MCC, Inman said he was committed to having “only top-notch people. There was to be no compromise on talent.”

So MCC began recruiting from outside the ranks of its member companies, a process that netted some highly respected researchers, but was also time-consuming, Smidt said. Delays in hiring were compounded in the software program, which started several months later than the others.

MCC has exceeded its initial employment projections. It now has 454 researchers and staff members. Since late August, some MCC staffers have been moving into the new facility, but even it isn’t big enough to hold MCC’s expanded staff; the software and semiconductor packaging programs will remain at other leased sites.

Staffing and space weren’t the only source of headaches. “There were loud disagreements, very strong disagreements” about technology transfer and licensing, said Inman, who said he took some heat from member companies for “being very honest about the problems--after they’ve been solved.”

Additionally, inside MCC, the researchers must wrestle with a cumbersome layer of bureaucracy. Only shareholder companies that participate in a given program (and pay the $1-million annual fee per program) are entitled to the results of that program’s research. That means researchers, whose work often could overlap from one program to another, are restricted from free exchanges of information.

Only three of the companies, Control Data, Harris and Sperry, participate in all four of the major programs. For instance, for a company participating only in the software program to take advantage of technology coming out of the database program, it must enter into a cross-licensing arrangement.

“MCC may be constrained because some companies are only funding parts. I would suspect it impedes their work, but it may be a necessary evil,” said Kenneth Flamm, a research associate at Brookings Institution specializing in foreign policy and technology.

The program directors admit that MCC’s structure is troublesome. And Inman said the company probably would not have attracted as many new members if it had required all shareholders to join each program. But, he said, he tells those who seek his advice in setting up new research consortiums to have each member participate in all programs “if they can do it.”

The consortium also will face more changes in membership. When Carrollton, Tex.-based Mostek was purchased by a French company, its membership was in doubt because MCC, to date, is not open to foreign interests. However, Westinghouse bought the division of Mostek involved with the MCC work and continued its membership. This coming year, BMC Industries will withdraw because it is divesting some businesses.

Two MCC members, Sperry and RCA, have been taken over by other companies and their memberships may not be continued by their new parent companies, Burroughs and GE respectively, beyond 1987 (a year’s notice is necessary to withdraw).

High turnover in memberships could cripple the consortium. Too many departures would drain the programs’ budgets; additions become more cumbersome the further the work progresses.

Inman said MCC also must begin branching into government research, make decisions about which research topics within a given project to pursue or drop and find programs to carry the project even further into the future.

Others will be watching as MCC goes about resolving those issues. Reporters are still streaming to Austin, eager to learn why it was chosen and to detail just what this “noble experiment” was producing.

Too, representatives from other high-tech companies and industries interested in beginning or restructuring their research ventures under the umbrella of the Cooperative Research Act, have been coming to MCC for clues.

Said Brookings’ Flamm: “It’s important we explore (what’s going on there), so we can learn from MCC how to make our national R&D; effort more successful.”