#Atari16 SIG
13 messages in this thread
A dumb question: why is it called Atari16 when the ST uses a 68000? Don't you
dare tell me to go ask them! —Mike
Depends on how you look at things Mike…
The marketing people at IBM always called their little beast a 16 bitter,
even though most of us were aware that it was an Intel kludge consisting of an
array of 4004s, communicating on an 8 bit bus, and doing most things internally
with 16 bits.
The 68000 is properly called a 16/32 bit processor, that way most people know
exactly what you mean. Of course it gives the marketing types nightmares.
Regards, Larry.
I think the fact that it operates on 32 bits is much more imprtant than the
fact it has 16 lines. As a programmer, I just want to know the bit width for
manipulation, and the speed of each instruction (regardless of what mechanism
is used to fetch and store data). I would say the 68000 is 32 bit like the
68020, but is twice as slow (or whatever) as the 68020.
Obviously I'm not a marketing type! —Mike P.S. As an electronics engr I
am interested in bus width, but it shouldn;t be important to the end user.
Read Motorola's official description of the 68000 — they describe it as it
really is — a 16 bit CPU.
Actually it depends on where you read the description. The Programmers
reference manual refers to it as a 16/32 bit micro.
Regards, Larry.
I don't have the Motorola publication (that I was quoting from) in front of me
now, so since I give give a page and paragraphic number — never mind.
Mike,
While I can see your point about looking at it from an "end user" point of
view, I still maintain that it is a 16/32 bit processor, and should be called
as such. My reasoning is that there is a considerable difference in speed
between the 16/32 and 32 bit versions of the same family of chip. This
difference is not only of interest to the engineer, but also to the end user
(provided he is using the chip, as oppoesed to s high level language interface
to same).
Regards, Larry.
What you say is true, but as you say, the only difference to the end-user is
speed. Using your (& everyone elses) nomenclature, the user needs to know
three things: the internal data size, the bus size, and the clock speed.
My point is that it would be simpler just to specify the data size and the
instruction speed, the latter based on bus size & clock speed, among many other
things unimportant to the end user.
I know I'm full of sh*t, but I don't wanna hear any more about it. It's a
sixteen bit machine, which means it should cost only two dollars. —Mike
Mike,
Well, technicaly, the 68000 is a 16 bit processor as it can only fetch
16 bits of data at a time, even if it uses 32 bit registers internally for
everything else and a 24bit address path. Motorola's official designation for
it is a 16/32 bit processor.
Don
Don,
See my reply to Larry's reply on this subject, which should be a few messages
prior to this one. It processes 32 bits at a time, even though it fetches
whatever, so I would call it a 32 bit processor. So there <grin> —Mike
Mike,
Alright, let's start up the war again<GRIN>. It depends on what you
consider it's capabilities–internal or external. I consider the amount of
data that can be fetched with 1 instruction to be the measure of bit size.
Others consider the amount of data that can be manipulated internaly to be the
bit size. Motorola chickened out and called it by both names.
Don
The 68000 is a 16 bit CPU just like the 8088 is an 8 bit CPU.