>>And let's not forget sir Clive Sinclair. He delivered 2.5 million (or more - that's just the last number I remember) cheap computers to kids who'd never have them otherwise. And he proved computers can be made cheaper.
>
>He also contributed one of my favorite quotes in the history of computing. One of the Sinclair PCs was delivered right around the time IBM shipped its first PC, the IBM PC being notable among other reasons for being the first personal computer with a 16 bit chip. Sinclair was asked why he chose to stick with an 8 bit processor. "Because I couldn't find a 4 bit chip I liked," he said.

Sounds like him :). He set out to make an affordable computer anyone can buy, and went down the fierce economy path: I think he's the first one with an ULA chip, what I see on most of economy motherboards today: just two chips onboard, one is the processor, and the other one does the rest. Everything else on the motherboard is peripheral - memory, video, disk etc.

The other trick was that his machine actually used faulty memory - he found a huge supply of cheap 4K chips, which were failed 8K chips where one side didn't work. After a while I figure he ran out of those and had to start using real 8K chips, which weren't quite used at all until Spectrum II (i.e. they were only half used), which then pretty much didn't need much of a change in wiring to address 128K of memory.

And, BTW, I don't think 8086 counts as a real 16-bit chip. It was, just like 8080 and Z80, an 8-bi chip with some 16-bit registers, and 16-bit memory addressing. It went beyond 16-bit memory by extra juggling - remember memory segments? I think even 80186 wasn't quite 16-bit. The "86" suffix, AFAIK, means "-8 and -16". The first truly 16 bit thing was the 80286.