In November of 1971 I was keynote speaker at the first Munich "SYSTEM" conference, now a binnial fixture. In the talk and in an exciting press conference that followed, I made a series of predictions about the disappearance of mainframe IBM competition vorldwide, during the next decade. As Munich is Siemens country, and I had some painfuI things to say about the parallels with General Electric in the States, there was quite a furor.

Since then I've used the material orer in talks and written pieces, with updates. The original pitch was made more than half a (hardware) generation ago, and since IBM is girding for the coup de grace, it looks like time to do a recap and survey the current position.

First, though, two definitions. By "computer company" I mean an outfit making mainframe hardware capable of imposing system architecture on a significant number of user data banks, data networks or number crunchers. "Significant" means, say, 1% of the world market, and draws the attention of the plug-to-plug boys and the software suppliers.

By "disappear," I mean either explicit or covert but obvious-to-customers withdrawal from the marketplace. There may be an announcement of new directions, a sale of important facilities or installed-machine base, a merger with another company. Above all, major development of the next generation of equipment is halted.

Now, which computer companies have disappeared, and how do the survivors stand? Chronologically, in order of my Munich dates, the picture is as follows AEG/Telefunken was to succomb in 1972, and was in fact forcibly merged with Nixdorf early that year. In 1974, after Nixdorf losses from the numbercruncher business became insupportable, Siemens was "asked" to take the Telefunken problem over, by the German government.

For 1973, I predicted the departure of Xerox, Philips and National Cash - the latter with the footnote, "an easig away; a withdrawal to the point-of-sale and cash register business." The Philips case is accurate; in or about 1973 it pulled back from the mainframe area and now does accounting machines, terminals and the bottom end of the Unidata spectrum (about which, more subsequently). Xerox (XDS) continues to announce new modifications of its beginning-of-the-decade line, but in terms of my definition, especially as regards a major next-generation super-storage, ultrachip development, has given up.

Whether this occurred in l973 or in 1972 or l974, an outsider can't say; it still takes orders still services its installed customer base. But, except as an expensive hobby for the parent Xerox management - or a face saver - I claim it has evaporated.

NCR is much healthier today under new top management than in 1971. The company still turns out Century systems and announces low-end improvements.

And it is venturing into terminals, intelligent and not so, as well as specialized point-of-sale store equipment. By my definition, though, and especially in terms of whether to expect a billion-dollar radically different mainframe system in the 1976-78 period, I think I can say National Cash has "disappeared" - probably more in l974 than in 1973, to use the Cheshire Cat model.

Dr. Herbert Grosch ist Editorial Director der Computerworld