6502/Z80 speed comparison (was MITS 2SIO serial chip?)

From: Richard Erlacher <edick_at_idcomm.com>
Date: Fri Dec 21 23:11:31 2001

I didn't bring those up, Allison. You're the one who brought up the 10 MHz part
and made the ill-considered statement that there exists any basis for comparison
between a Z80 and a 6502 at the same clock rate, or even with the Z80 at twice
the clock rate. A 10 MHz part didn't exist back then either, nor did the latest
shrink of the CMOS 6502. What I've been going on about is a way of reconciling
the major architectural differences that make a comparison on the basis of
performance of the processor core/instruction set possible. Talking simply
about clock rate is almost meaningless, since the way in which the processors
behave with respect to the clock differes greatly.

If you want to insist that you can devise a task, any task at all, mind you,
that you can code in Z80 code to run at 2x the clock rate of a 6502, in, say,
<80 lines of assembler, that's strictly a computing task, just to leave I/O out
of it, or any I/O task that you think you can code in legitimate Z80
instructions, providing a precise spec for the I/O task, then I'd certainly come
out and say you can't cook a legitimate one up that a 6502 can't accomplish in
less time, in the case of the computing task, and that, since the I/O task spec
limits the rate, the 6502 will be able to do it as well. I'd like to see a task
that meets those spec's.

I'm pretty sure that a 2 MHz 6502 can easily handle transfers to/from an 8"
floppy drive via a WD-179x using MFM without any hardware synchronization using,
say, the wait line. I'm also pretty comfortable that the Z80 would have to
strain. I certainly don't know that it's impossible, but every controller I've
seen that handles MFM on 8" floppy drives for use with a Z80 uses the Wait line
to sync up with the DRQ signal. For a 2 MHz 6502, it's a leisurely task
requiring no hardware for synchronizing the process at all.

The crux of making a reasonable comparison between a 6502, avaialble in the
'70's, and a Z80, also available in the '70's, is selecting a clock and hardware
arrangement that makes it possible for the Z80 to compete with the 6502 on a
fair basis at all. To run without waits, the Z80 requires relatively fast
memory, which, in its day, would make the system too costly. Running the Z80
with the same memory that the 6502 requires would make the Z80 too slow to


----- Original Message -----
From: "ajp166" <ajp166_at_bellatlantic.net>
To: <classiccmp_at_classiccmp.org>
Sent: Friday, December 21, 2001 7:27 PM
Subject: Re: 6502/Z80 speed comparison (was MITS 2SIO serial chip?)

> When you get over it...
> Talking about 20mhz 65C02s and other "fast" parts that didn't exist when
> even the z180 (orginally 64180) was introduced is equally bogus.
You brought 'em up, I didn't.

My main 6502, back in '79, was a Synertek NMOS 6502C, specified to run at a 4
MHz clock. They were production parts, avaiable through every one of their
distributors. They stopped making those when the CMOS part became available,
since it was more profitable.
> A Pentuim anything running a Z80 emulator is still emulation and not
> native silicon.
A Z180 is not a Z80.
> Now cut the crap.
> Allison
> -----Original Message-----
> From: Richard Erlacher <edick_at_idcomm.com>
> To: classiccmp_at_classiccmp.org <classiccmp_at_classiccmp.org>
> Date: Friday, December 21, 2001 7:06 PM
> Subject: Re: 6502/Z80 speed comparison (was MITS 2SIO serial chip?)
> >We went through all this a couple of years back under the same heading.
> If the
> >Z235980 at 234 THz is code compatible then, if I understand you
> correctly, it IS
> >a Z80, right? Even though the Z180 won't fit in a Z80 socket, you'd
> sell it as
> >a Z80 anyway, right? Even though it didn't even exist back when it
> mattered,
> >you still insist it's a Z80, right?
> >
> >My Pentium executes the Z80 code just fine at about 75x the speed of of
> a Z80.
> >Does that mean it's a Z-80?
> >
> >We're comparing CHIPS, not philosophical constructs. If it IS a Z80, or
> Mostek
> >3480, or something else EXACTLY a Z80, i.e. built under the license,
> >pin-compatible, code-compatible, etc. then MAYBE it's germane to this
> >discussion. No chip that isn't a pin-compatible substitute commonly
> referred to
> >as a Z80 back in the days when the Z80 mattered is germane to this
> topic. If it
> >won't plug into the socket of a Z80, FORGET IT, because it's not a Z80.
> If
> >that's too difficult for you, then please ask an adult why a 47-ohm
> resistor
> >isn't the same thing as a 75-ohm resistor.
> >
> >I'm sure glad you're not trying to sell parts any more, Allison. I'd
> hate to
> >have to argue with you that the choke you're trying to pitch isn't a
> diode.
> >
> >Dick
> >
> >----- Original Message -----
> >From: "ajp166" <ajp166_at_bellatlantic.net>
> >To: <classiccmp_at_classiccmp.org>
> >Sent: Friday, December 21, 2001 4:24 PM
> >Subject: Re: 6502/Z80 speed comparison (was MITS 2SIO serial chip?)
> >
> >
> >> From: Richard Erlacher <edick_at_idcomm.com>
> >>
> >>
> >> >There are lots of things that you could compare, but the first things
> >> you've got
> >> >to leave out are the ones that aren't a Z80, which immediately
> deletes
> >> the Z180,
> >> >and Z280. The Z80 is not around any more than the 6502 is around.
> >> There are
> >>
> >>
> >> Why? they are still z80 core and code compatable. While they add
> things
> >> like
> >> serial IO, timers and MMU they are Z80, maybe more so than 65C02.
> >>
> >> Allison
> >>
> >>
> >>
> >
> >
Received on Fri Dec 21 2001 - 23:11:31 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:33:41 BST