Motorola Inc came out swinging last week, trying to take the offensive against chip rivals Sun Microsystems Inc with the Sparc and MIPS Computer Systems Inc with the R-series, both of whom have been riding roughshod over Motorola for months now. Dubbing it the great RISC myth, Motorola contends that the battle for pride of place is far from over. In fact, if you follow its logic, the decision won’t be known until sometime in 1995 when the RISC market overtakes conventional complex instruction set microprocessors. It also contends that its chances of dominance are better than the others despite the fact that the competitors’ volumes are ahead at the moment and have been for a few years. Motorola has had one of those voluminous white papers done up by 88open and with 88open officials in tow was making the rounds of press folk last week giving reporters a peek at their view of the world. Motorola and 88open say that the weight of the 400 engineers it has working on the 88000 and its $500m investment in a new manufacturing facility to fabricate 88000 chips until the end of the decade has to count for a lot, especially as the leading edge moves from 1 micron technology to 0.5 and smaller. Its resources alone guarantee a long lifecycle and several generations of the 88000, with the company planning to rev the thing to a single 300MHz 110m-device chip by the late 1990s. Motorola charges that the small competiting Sparc semiconductor makers – and MIPS’ too – don’t have the financial clout to move to 0.5 micron without considerable outside assistance – although Cypress Semiconductor Inc with its X-ray stepper on the way would beg to differ (CI No 1,667). And the multiple as opposed to second sourcing practices employed by Sun and MIPS means that quality can vary widely, and that compatibility is unassured. Where Motorola tries to score the most points is targeting Sun’s business plan. It claims Sun’s clone approach leaves nowhere for its design wins to go. Sparc clones have to compete on price or in market segments that Sun has yet to enter. But Sun continues to drive prices down through increased manufacturing and economies of scale. It also plans to move into the few remaining market segments in which it doesn’t already compete, such as multi-processors and portables. Unlike the personal computer model, Unix system builders have to pay royalties and licensing fees for proprietary technology and, in addition to the heavy technical support required, there is no high volume commodity market driving it. In picking Sun apart, Motorola also claims than Sun has no real common Applications Binary Interface for shrink-wrapped applications portability across Sparc-based systems, creating a crisis of confidence for independent software vendors that are unlikely to support their software on anything but a Sun system. And despite Sun’s recent division of the company into three entities, Motorola still wonders whether Sun will develop the Sparc to its own best advantage, and use early access to gain market advantage over the cloners. Intel Corp is dismissed out of hand as unlikely to devote the resources to get the 80860 widely accepted, its only ready market being as a graphics co-processor.
67 design wins
The independence of its supporters club, Mass860, which works out of the Intel office, was also questioned. MIPS licensees and OEM customers, Motorola claims, have few if any new market opportunities because they are going directly after Sun and face tough competition and pricing pressures. MIPS’ position on Application Binary Interfaces is worse than Sun’s, Motorola says. Its design wins aren’t a community but rather a series of incompatible architectures. And as far as the Advanced Computing Environment goes, no one, not even the ACE members, know if it will be successful in providing software. Motorola claims for itself the only real Applications Binary Interface and the first real shrink-wrapped applications to exist in the Unix industry. More than 150 applications, it says, are now certified as running on any 88000 implementatio
n. And the designability of the 88000 gives it design wins the freedom to pursue unique and emerging markets as well as a quicker time to market. The high-performance large-scale network server market is wide open to them. 88000 vendors are not competing on price. Motorola claims 67 design wins, 32 in computer systems and 35 in embedded applications. Only 23 systems makers are announced and shipping today. Four are publicly committed but unannounced. Seven are privately committed and unannounced. So how come Motorola lags the pack? – Maureen O’Gara
