My inspirator has been Douglas W. Jones
[here]
and his artile "The Ultimate RISC"
[here].
[``The Ultimate RISC,'' Computer Architecture News, 16, 3 (June 1988) 48-55.]
I Created some neat small ZISC simulators, but discontinued (lack of resources).
Noticed that IBM has registered ZISC as trademark - but that is for totally different technology...
Also noticed that Guy Paillet has patented
ZISC tehnology (which really is a neuron technology).
By ZISC I do not mean neuron or neural net computer, this ZISC architecture is a state machine
(as presented during the 80's). Naturally it is possible to create "neuron cells" using a very
fast state machine like ZISC.
Building fast ZISC machine has been my dream for about 15 years, but until now it has not been possible,
because ZISC uses large amount fast memory.
Good points in ZISC design:
possibility to build really large scale parallel system - if building multiple ZISC
components in one chip is possible.
highly optimized hardware.
possibility to emulate existing assembly, but naturally only within one ZISC element.
