z80 timing... 6502 timing

From: Richard Erlacher <edick_at_idcomm.com>
Date: Mon Apr 19 18:42:06 1999

While what you say is certainly true, the page-zero usage is not dynamic,
so, unless your task is very large and complex, page zero usage is therefore
under the coder's control. I agree with your observation that the rate of
memory usage is a good indicator of the rate at which a processor gets work
done. The 6502 shines in that respect, in that the internal operations are
generally overalpped with the fetch of the next instruction, with a few
notable exceptions.

Generally speaking, I've chosen processors with foreknowledge of what the
device's most frequent tasks would be. For example, I picked a 65C02 for an
application for which my boss had expressed a profound preference for the
Z-80. I had to prove to him that, for OUR task, the 65C02 had a couple of
features we could use very effectively and which the Z-80 didn't offer.
It's always case specific.

Dick
-----Original Message-----
From: Dwight Elvey <elvey_at_hal.com>
To: Discussion re-collecting of classic computers
<classiccmp_at_u.washington.edu>
Date: Monday, April 19, 1999 2:55 PM
Subject: Re[2]: z80 timing... 6502 timing


>"Richard Erlacher" <edick_at_idcomm.com> wrote:
>> I don't know why this has to be so complicated. There need to be
>> constraints in order to ensure a level playing field, but since there are
>> two related objectives, (1) to find out which of the two processors in
the
>> title of this message is "faster" and (2) to generate the fastest code
for
>> them for comparison.
>
>Hi
> I always have to wonder about the meaning of faster.
>Although I like the 6502 for a lot of things and would
>say that it is often faster than a 8080 for many things,
>I also know that resources like page zero get used up
>quickly. Once these are gone, things tend to slow down.
> I think we will continue to see that, like in the past,
>most processors with about the same memory access speeds
>will do about the same amount of work ( same bit count ).
>On chip cache has changed that some but that just makes
>another category. I would say any processor was superior
>to another on small benchmarks unless the numbers were
>in the 5X magnitude or more. Large application are more
>useful in comparisons but it is hard to come up with
>enough cross platform examples to make meaningful judgments.
> I've always said, one should stick with what makes you happy.
>Dwight
>
Received on Mon Apr 19 1999 - 18:42:06 BST