The withdrawal of Control Data Corp from the supercomputer business – which it effectively invented, and managed to lead until Cray Research Inc came on the scene, much to the chagrin of IBM – should have set alarm bells ringing insistently in Washington, but the devastating message isn’t really getting through. Even the announcement from Cray Research Inc that it could no longer afford to fund two families of processors and was sending Seymour Cray off to play in a shiny new cave in Colorado Springs – in the premises once occupied by Inmos International Inc – hasn’t led many to reach the inescapable conclusion: America’s electronics industry is not hopelessly ill-structured to remain competitive with Japan Inc in the medium term. Control Data Corp is a critically weak company, quite unfit to be one of the only two US standard bearers of a leading-edge technology; Cray Research, with annual sales still well shy of even one billion dollars – smaller even than Data General Corp for heaven’s sake, might have been an appropriate champion when the only competition came from venture-capital funded hopefuls in the US, but in an age when Japan Inc has identified supercomputers as a crucial technology, a Cray is woefully ill-equipped to hold the Alamo. Half pityingly Commenting on the demise of Control Data’s ETA Systems, a Hitachi official declared half-pityingly that his company’s effort in supercomputers was costing Hitachi $100m a year – but that was a price that had to be paid: supercomputers were an area where Hitachi had to be a player. So tiny little Cray Research, 5% smaller than it was a month ago after kissing goodbye to Seymour, is lined up against not one but three Japanese supercomputer manufacturers, the smallest of which is a $17,000m a year company (Fujitsu Ltd hasn’t announced its latest consolidated figures yet, but they are likely to be around that level – unconsolidated turnover was $14,000m, so Fujitsu, not DEC, is emphatically the second biggest computer company in the world after IBM). NEC Corp and Hitachi Ltd are each somewhere around half the size of IBM, but each has substantial interests outside computers, telecommunications and semiconductors. And that is the point. Of the US chipmakers, only Texas Instruments Inc and Motorola Inc come anywhere near an appropriate profile to compete with the Japanese: Intel Corp, at only $2,000m a year, almost all of it from semiconductors, is ludicrously small to try to compete with companies 10 times that size. The companies that have so signally failed America are the counterparts to the Japanese giants – IBM, although it is belatedly beginning to wake up to its responsibilities to its homeland, General Electric Co, Westinghouse Electric, United Technologies. Each of those companies has long been big enough to stand up to the Japanese electrical and electronic congolmerates – but instead of fostering key technologies and market areas, they have at each setback retreated, abandoning their interests in semiconductors, computers, consumer electronics, robotics. Today each is clearly too fat and comfortable to want to face up to the real challenge of the 1990s – so where can America now turn? Busted flush Michael Blumenthal was absolutely right when he decided that an order of magnitude leap in size was needed to compete effectively with IBM and Japan Inc, and that Burroughs was much too small to go it alone, but sadly he still seems to think that computers are just like any other commodity, so that in each of the deals he has struck to create Unisys Corp has resulted in one and one adding up to a lot less than two. Instead of going after another busted flush, Sperry Corp, the appropriate move for Burroughs would have been to acquire a company of the same size that was in a related area – Xerox Corp is an obvious example, but Motorola Inc would have made sense too – preferably both, over time. And then Zenith Data Systems instead of Convergent Inc. At that point, he would have created his dream of putting together a $20,000m a year company, but one with a broad enough
base that when one part o fthe business – such as chips – went into a downturn, others were there to ensure that the figures weren’t too awful, and the chip side, instead of making cuts, was able to trade through its difficulties. That is why the idea of a merger between Cray Research and Apple Computer Inc was not as crazy as it sounded at first hearing – but such a merger would be a very partial solution on its own. To such a combination needs to be added DEC to fill out the middle ground, plus a chipmaker Motorola again best fills the bill, and brings with it a welcome diversification into data communications products. At that point, a company is created that has the critical mass to stand head-to-head with the Japanese. A similar combination of Intel Corp and Compaq Computer Corp with either Hewlett-Packard Co or NCR Corp, plus two or three more complementary businesses would leave the US with three companies capable of taking on Japan Inc on equal terms. The problem of course is that so many of companies named are still run by their founders, who are loth to throw in their lot to build something that would soon look very different from their creation. Moreover the wheels of American ingenuity are fuelled by stock options that need to be tied directly to the performance of one part of such a conglomerate. The only solution would then seem to be alliances rather than full mergers, – but with no more than 25% or so of the equity of an acquired company remaining out on the exchanges. Half-hearted It looked like a half-hearted effort by IBM to bankroll Steve Chen’s Supercomputer Systems Inc instead of establishing its own supercomputer development effort, but given the American entrepreneurial psyche, it might well prove to have been the most appropriate solution. The other weakness of American companies compared with their Japanese – or German – competitors is the absence of friendly banks that will finance them forever if need be, and friendly shareholders for whom the long term is 20 or 30 years. But the latter may have arrived – at least in the case of IBM: the company’s shares now stand just shy of $110 – less than 10% above the low they made in the wake of Meltdown Monday, yet there is no talk of institutions managing the mites of widows and orphans getting restive, no talk of a break-up bid to maximise shareholder value: perhaps IBM has discovered a whole new breed of American shareholder blessed with the patience of a Deutsche Bank or a Nomura Securities, in which case all that is needed to revitalise US industry is two or three megalomaniacs capable of infusing the likes of Ken Olsen and Gordon Moore with their vision.
