JANUARY 1984 - VOLUME 5 - NUMBER 1
Dreaming Up the Future Computer
Facing tough Japanese competition, American firms are teaming up to develop the next generation of computersby Mark Clifford
"It's time to play hardball with Japan."
That's the battle slogan of Control Data Corporation chairman William Norris, who has spearheaded a newly formed research consortium comprised of 13 major U.S. computer and semi-conductor manufacturers.
Over the past two years, Norris has emerged as the leader of a counterattack by U.S. high-tech executives against the Japanese electronics industry and what the} consider Japan's "unfair" national policy of targeting specific industries for rapid growth. While Norris's more controversial proposals to ban Japanese graduate students from American universities have flopped, the fledgling Microelectronics and Computer Technology Corporation (MCC) promises to be a more significant force in the area of high technology politics.
Former National Security Agency director Bobby Inman heads the consortium, which plans to set up its headquarters in Austin, Texas. With a projected annual budget of $50 million to $100 million, the joint venture will conduct long-term basic research in computer and software technology. In-addition to Control Data, MCC members to date include Motorola, Advanced Micro Devices, Allied, Digital Equipment, Harris, Honeywell, Martin Marietta, Mostek, National Semiconductor, NCR, RCA, and Sperry. Annual electronics sales of the 13 firms total about $32 billion.
The need to cooperate, Control Data's Norris said last January, "stems principally from the Japanese threat. If the U.S. does not respond successfully, it will suffer irreparable losses in both domestic and international markets."
In a speech to two dozen top executives the following month, Norris said, "This threat from overseas is compounded by the fact that micro-electronics and computer firms in the United States are suffering from a combination of scarce capital resources, high capital costs, and a rapidly growing need for capital to develop and exploit appropriate new technological possibilities."
Like Norris, Motorola Corporation chairman Robert Galvin thinks it's time to get tough. "The continuation of the status quo, where the practices of the other side are so much of an anathema to our system and our country, is intolerable if we wish to preserve the American system," he said in a speech last year.
The Semiconductor Wars
According to industry analysts, Japan's production of semiconductors presents a direct and serious challenge to the U.S. industry. The Semiconductor Industries Association has reported that a cycle of disinvestment is taking place in the U.S., with manufacturers generating insufficient profits from the current generation of products to fund research on the next. It noted that 15 U.S. manufacturers produced the previous generation of memory chips, the 16K RAM, while only Motorola, Intel, and Mostek produced a significant volume of the latest generation of semiconductor memory products, the so-called 64K RAM chips. But these three companies suffered a combined loss of $46.5 million on these products in 1982, a loss the Semiconductor Industries Association attributes to Japanese competition.
Significantly, the Japanese got their first major opening in the U.S. semiconductor market during the late 1970s and early 1980s, when U.S. manufacturers were unable to keep up with demand for 64K RAM chips. Productive capacity in the U.S. was insufficient because U.S. companies had slowed down on investments during the 1974-75 recession while the Japanese had continued to invest. The ensuing void in the American market was quickly filled by Japanese companies.
But U.S. executives aren't simply worried about semiconductors. Japanese manufacturers have already captured a significant domestic market share for some of the lower-priced computer equipment, such as display terminals and printers. More ominous, from the U.S. corporate viewpoint, are Japan's intentions to move into the mainframe computer field-the giant computers that are used almost exclusively by multinational corporations and large institutions such as the Pentagon.
Under the leadership of the Japanese Ministry of International Trade and Industry (MITI), eight Japanese computer manufacturers have joined the semiofficial Nippon Telegraph and Telephone corporation (NTT) in a major research and development project to develop a fifth generation of computers. The $450 million project is being funded partially by MITI, but most of the funds will come from the companies themselves.* Among the goals of the project are the development of computers that can understand the human voice, and "supercomputers," which have previously been the exclusive domain of two U.S. firms, Control Data and Cray Research. Global competition for these new models over the next ten years is expected to be stiff, and both the U.S. government and American companies are afraid they may lose the contest.
MCC's research will concentrate on the areas that the Japanese companies and government have targeted. The cornerstone of MCC's agenda is its AlphaOmega project which will develop advanced computer programming for the next ten years. In addition to working on the computing functions of the fifth generation computer, the Alpha-Omega project will study artificial intelligence problems and applications.
A seven-year software productivity project will concentrate on new tools for software development while an eight year program will focus on "CAD/CAM," or computer-aided design and computeraided manufacturing. The CAD/CAM project will develop the design tools necessary to construct the silicon chips for the fifth generation computer. A fourth program will, over a six-year period, develop packaging and automatic insertion methods for future generations of silicon chips.
In sum, MCC is attempting a program of vertical integration. It hopes to lay the groundwork for many of the components necessary to develop the generation of computers which will prevail a decade from now.
Challenge to Antitrust Laws
MCC sounds ambitious, and it is. But it poses an unprecedented challenge to antitrust laws, even if the consortium is designed to engage only in basic research.
The Reagan Justice Department gave an informal go-ahead to MCC last December, but that isn't enough for its backers, such as Control Data's Price, who complains that "the Justice Department seems to enjoy the uncertainty" of possible antitrust action.
MCC's supporters are currently trying to push legislation through Congress which would exempt joint research ventures such as MCC from the threat of antitrust action and the possibility of damages. Senator John Glenn is sponsoring one 4 the bills, the Antitrust Joint Research Act of 1983. Len Weiss, Glenn's science and technology advisor, says that joint R&D, ventures have become "a matter of survival" in the high-tech field. "We've gotten to the point where for companies to compete, on an international basis with companies abroad we've got to tighten up our operations here," Weiss said, adding that "the Japanese certainly allow such activities [joint ventures] to go on."
Weiss insisted, however, that Glenn's commitment to facilitating joint ventures "in no way means we are looking to relax our antitrust laws." He says that there is a clear line between basic research, which should have the potential to be shared, and applied research, which would remain subject to antitrust restrictions.
The Reagan Administration has jumped on the bandwagon as well. In September, the administration proposed regulations to "stimulate the creation and development of new technology, increase the country's productivity, and enable our industries to compete more effectively in world markets."
These bills are pending action in House and Senate committees. One congressional aide says that House action is not expected soon because Peter Rodino, the chairman of the Judiciary Committee, is "skeptical about the whole issue."
But MCC's relationship with one agency of the federal government-the Defense Department-remains unclear. Earlier this year the Defense Advanced Research Projects Agency (DARPA) announced the formation of a "Strategic Computing and Survivability" project to create a new generation of supercomputers. With a multimillion dollar budget, DARPA will sponsor research in both computer and semiconductor technology. MCC president Bobby Inman has admitted that the private and government agencies have similar goals and might be "competing for the same talent." However, as a former director of the National Security Agency and a high official of the CIA, Inman's job might be to find common ground between the government and the private sector.
And MCC has another problem: not all those it seeks to help are anxious to join. Technology leader Intel is absent from the MCC roster as is Texas Instruments. And the two largest companies in the industry, AT&T and IBM, have expressed no interest in MCC. (Although IBM reportedly declined to join for fear of provoking yet another round of antitrust litigation, a more likely scenario is that IBM had more to lose, in the way of trade secrets, than to gain.)
Indeed, MCC may be designed as much to counteract the strength of IBM and AT&T as the Japanese. A staff assistant to Senator Gary Hart (D-Colorado), a key player in high-tech politicking circles, noted that "the deregulation of the telecommunications industry and the decision of IBM to enter other segments of the market may also have stirred interest" among MCC backers.
IBM's revenues of nearly $29 billion from electronic sales alone nearly equal those of MCC's charter members combined. And even after divestiture, AT&T's sales will exceed that mark. These two firms alone have the capability to conduct research on a scale to compete with the nationally-coordinated efforts of the French and the Japanese.
Pacific Studies Center director Lenny Siegel, who describes the scope of IBM's research as comparable to that of Japan's, says that the alternative to MCC "would be survival of the fittest and a lot of smaller companies would go under." Siegel contends that rather than enhancing competition, antitrust laws in this instance would stifle it. "It's pretty clear that invocation of antitrust laws against MCC would lead to more concentration."
Author Yoshi Tsurumi, Professor of International Business at the City University of New York and a frequent commentator on industrial policies in the U.S. and Japan, agrees. Lifting antitrust restrictions on research is "pro-competitive," he says, and points to the example of Japan, where three or four competing firms "pull their resources together for basic research" but then "three or four" companies compete like hell when the research bears fruit. If you don't let potential candidates pool their resources, then a monopoly is created." Passing laws to facilitate joint efforts in research, he adds, "would bring the U.S. into the 20th century."
Tsurumi cautions against what he calls the "unfortunate characterization" of Japan as a country that lacks competitive industries. "The causes of falling behind in the U.S. are not due to antitrust constraints, but the short-sighted U.S. management attitude," he argues. "Japanese executives seek technical, not financial solutions. But when the Americans start losing out to the competition, they cry protection."
The critical question for MCC, he concludes, is "whether members are capable of running with the fruits of the research through the competitive commercialization process. That's where the Japanese strength lies."
* The companies involved in the project are: Fujitsu, Nippon Electric, Toshiba, Hitachi, Mitsubishi, Oki Electric, Matsushita, and Sharp.
Mark Clifford, formerly an associate editor of Electronic News, is a business and financial writer in New York.