24 Jul
2014
24 Jul
'14
3:21 p.m.
I also second the idea of having an idea what you want
to do with the
machine. That way, you can decide what kind of system to look for. A
I'll expend a little on that and give a personal view.
I am a hardware person, so yo me the attraction of a minicomputer is that
(at least i nthe machiens I like) the CPU is not one chip. It's lots of
ICs, often spread over several boards. So I can 'get inside' the CPU, I
can trace the microcode, watch data folowing between ALU and registers,
and so on. So _for me_ an LSI11 (or any of the later ASIC-based minis
from any manufactuer) is not really what I am looking
for.
YMMV, of course. which is qhy you must think of what you want ot do with
the machine, especially if you are paying real money for it.
For _my_ interest there are otehr machines that suit, although they are
not really minicomputers. The first is the HP9800 series of calculators
(HP9810, 9820, 9821, 9830). These use a 16 bit bit-serial processor,
built from about 80 TTL ICs. Although they are obviosuly a lot less
expandable than a minicomputer, and thye don't run complex operating
systems , etc, they are machines you can investigate tthe CPU operation
in.
The other machine is the ICL/3Rivers PERQ (if you are lucky enough to
find one!). This is a workstation. But the CPU is a board of over 250
ICs. Again, youc an 'get inside' it. This will run Unix, but is a lot
less expandable than a minicomputer. It also is one of the few machiens
you're liekly to see where hthe CPU microcode is loaded into RAM at
boot-up, and where the CPU is entirely gerneal-purpose [1] so you cna
redfine the insturcion set if you're so inclined.
[1] The VAX11/730 also loads the CPU microcode at power-up, but from what
I rememebr, ther eare some hardware features that optimise it for the VAX
instruction set. While it is possible to give it a differnet instruciton
set, I think it wopuld be a lot less efficient.
-tony