- Ya, I'm late to reply to this thread ...
I kinda remember the 70s and 80s when there were vague classes of
computers: minicomputers, midi and mainframe. I'd say most were
identified by their physical size and cost.
Mainframes are easy to identify: large cabinets in the air conditioned
"hands off" machine room. Costs millions of dollars. They still do.
Minicomputers start the size of a breadbox and grow into 1 rack, then
2 or more (such as the PDP8, PDP11, etc). They're known not so much
for their capabilities but their limitations: maximum amount of
addressing/RAM, how many users before it gets too laggy, etc.
Midicomputers were maxed out minicomputers, probably back when
"departmental shared-computers" were advocated as a decentralized
alternative to company-wide mainframes. I suspect large CAD/CAM
systems such as wall-sized GE Calma were considered midicomputers (it
filled the room with the dedicated color graphic terminals, digitizers
and large scale plotter). The image that comes to mind is a VAX
11/780: boxes too small to be called "mini".
I assert that the evolution of the microprocessor CPU is the main
driving force for the blurring of the classes. Once minicomputers
(even microcomputers) broke the 1 meg RAM barrier, they were getting
into the "big league" for solving problems that once required
mainframes. For a while, "workstations" were high-end vs.
Personal-computers, but with today's high end gaming systems with
graphics co-processors outrunning the CPU, most labels no longer apply.
- Yeep! Is there a way to edit postings here?
> Midicomputers ... The image that comes to mind is aMicro/mini/midi/mainframe is more of a "I know it when I see it" thing
> VAX 11/780: boxes too >> LARGE << to be called "mini".
than a list of parameters.