Talk:Motorola 88000
From Wikipedia, the free encyclopedia
In the late 1980s several companies were actively watching the 88000 for future use, including NeXT and Apple Computer, but both gave up by the time the 88110 was available in 1990"'
So why did they give up?
[edit] Darn good question
I used the 88k for a "high-powered" embedded system in the late 80s and I thought it was a really nice CPU, easy to design with, program for and easy to get good performance with minimal tweaking. At the time it had some very advanced concepts which made efficient pipelining a snap.
However, the multi-chip design of the 88100 (needed two MMUs and one CPU; remember, split I/D buses!) was a big turn-off in general, initial performance was disappointing, the separate MMU chips slowed things down even more, and it wasn't cheap. The 88110 was a major improvement in both speed and ease of use, but they were always behind in the CPU clock speed wars... and wasn't long after that PowerPC was announced. Combined with Sun's heroic efforts to spread the gospel of SPARC, the chip was doomed. A shame, but I think the answer is that just came out at exactly the wrong time. If the 88110 had been released in the late 80s, it probably would've caught on.
I believe Ford (yeah, the car people) also used it in a graphics workstation, but I may be misremembering. The 88k actually had special instructions for doing Gouraud shading calculations.
i have m88k in NCD MCX Xterminal, works nice, uses plain simm (72pin) as memory and so on, bootable via tftp, just stating that more than Apple were using this cpu.