killing the x86

Jecel Assumpcao Jr. wrote: The 960 had already lost all hope of being a general-purpose processor by that time. The 960 was the BiiN/P7 processor, with the tag bit and object-oriented operations desupported (and removed in some later versions). It was originally intended to be the replacement of the iAPX 432, in that it was an attempt to take the good parts of the 432 but ditch the complexities of the CISC architecture that were viewed as being a substantial part of the performance problem. Rather than going it alone, Intel partnered with Siemens, and the joint venture became BiiN. It was a market failure, and the 960 came about as Intel tried to salvage something from it. There was no significant push to make the 960 a general purpose processor, as that ship had already sunk. Whether Intel *wanted* to evolve the 486 or the 860 is irrelevant. They needed to make what customers wanted (and thus were willing to purchase), and 99.99% of customers wanted a faster x86. That affected the relative engineering resources that could be put into the followons for the 486 and 860. It's similar to the reason that Intel did NOT get clobbered in the DRAM business; without having an upper management edict to exit DRAM, they had already effectively done so, because wafer starts were allocated to products based on the profitability of those products. (The previous sentence is the only one that I can back up Survive_.) Eric