> > Shure ? I had to go thru the same kind of decision, and I
> > learned to love the 186 - shure, one needs a little push
> > to start, but later on it's just great - eventualy the best
> > _sixteen_ bit CPU around. If you look at the 8080/5, Z80,
> > 8086 and 80186 family, the 186 is the finest of all.
> I would probably prefer the V30 which had the 186 instruction
> set, but the 8086 pinout.
minor details - just I liked the on chip components a lot.
The CS logic can be used under special circumstances almost
like a MMU, and the DMA chanels are very handy (and usable
unlike the PC style DMA). You may have up to 3 I/O operations
running at the same time - two via DMA and one by INS/OUTS.
And if your application needs to shovel memory around within
a tight time limit, the DMA will do the job faster than the CPU.
> > Sleak and simple - compared the 68K looks quite bloaded
> > and clumpsy (ok, it's 32 bit, but still the most compared
> > competition at it's time).
> Do I hear another war starting? According to most sane people,
> and Motorola, the 68K is a 16 bit processor.
Bait Taken.
One has to select his favorate definition, and I selected the
largest data size supported by most (all) basic instructions.
In case of the 8080 it's 8 Bit (even there are some 16 bit
instructions), for the 186 its 16 bits, and for the Motorola
it's 32.
(It has to be added here that we're only talking about basic
instructions on GP CPUs - no FP, no vectorandwahtever data)
There are different measurements around, like data path size,
register size or _marketing_ size.
Now let us look at the 68K family
68000:
Data path: 16
Register: 32
ISA: 32
Marketing: 16
So if this is a 16 Bit CPU, what's a 68008 ? 8 Bit ?
Still the same CPU core - and further more, is a 68020
now 32 Bit ? Again, wheres the 'new' 32 Bit part?
Hard to see a fundamental Difference. Ob, even further,
take the NS 16032 vs NS 32032 enhancment - a 16 Bit CPU
(with 32 Bit bus) versus a 32 Bit CPU - isn't it ?
Rubbish - both are exactly the same chip - just marketing
department decided that it's time to hit the new emerging
32 Bit market. Around 1978 when NS and Mot did come up with
their 32 Bit design everybody was asking (hypeing?) for
16 bit CPUs (instead of 8 bit ones), so marketing named the
stuff the 'sellable' way (Remember, Intel did just around
this time a grande flop with their i432 32 bit CPU).
Well, this part is marketing, you still may argue that a
external data bus is a valid scheme for naming - nice, but
then every x86 chip since the original Pentium is a true
64 Bit CPU. Sounds wrong, huh ? But it's true! (according
to said scheme)
During my practice with several generations of /370ish
mainframes I learned that there is basicly no measurement
based on physical aperance that cant be twisted. So the ISA
is (for meone :) the only measurement with _some_ creditibility.
/360&up-ish CPUs where available with ALUs from 8 to 32 bits,
internal data pathes from 8 to 256 bits, external data pathers
from 8 to 1024 bits (128 bytes per memory cycle), but nobody
will doubt that they are all 32 Bit CPUs.
Next in line.
:)
Gruss
H.
P.S.: And when you think a bit further you'll notice that the
physical implementation of an ISA doesn't matter at all - the
only relevance is the resulting speed - and here the inpact of
some more megabytes of cache can be more important then buswidth,
ALU size or whatever. Eventualy a turing machine may wind its
tape below your beautiful 64 bit ISA, without any implications
for your task.
--
VCF Europa 2.0 am 28./29. April 2001 in Muenchen
http://www.vintage.org/vcfe
http://www.homecomputer.de/vcfe
Received on Tue Nov 21 2000 - 11:42:15 GMT