On Sep 4, 11:37 am, "Michael J. Mahon" <mjma...@aol.com> wrote:
Or, you could do what may be the best test: running a couple
of real applications that have some relevance and measuring the
speedup on those--an Appleworks sort or text search would be one
possibility. I'd use one of my programs that runs long enough
that I have to wait for it to finish...
The problem, of course, is broad comparability and verifiability.
If you run a popular benchmark, it makes it easier to compare Apples
to "oranges", but at the cost of realism. If you run a real application
that is meaningful and "real", the results can only be compared for the
machines and configurations that you, yourself, measure.
Computer performance measurement is a many splendored thing, and it
steadfastly refuses to be characterized by a single number--even for
our relatively simple older machines.
I completely agree that a measurement of actual speedup in
applications you actually use is the best measure of any performance
improving hardware addition.
I would add though that in my experience, the most representative
synthetic benchmark (for the given case of same machine architecture,
different clock/cache configuration) is memory bandwidth.
In the case of the Apple II, I think it's fairly safe to assume that
compute-heavy tasks will show a speedup proportional to clock rate;
it's not like we have to deal with ISA compatible chips with wildly
different pipeline depth, branch prediction and cache behavior like is
common in the Intel/AMD world.
Yourself and I conducted a very informal (and contrived) test a while
back (posts archived on google groups) that showed a very large
difference between memory write bandwidth between aTranswarpI and aZIPthat was completely disproportionate to clock speed.
Subjectively speaking, aTranswarpI and a 4MhzZIPfeel pretty
similar when doing some tasks (ie full screen editing inMerlin).
Assembling a large program however shows a marked difference between
the two.
It's been my general experience with other systems too, that with the
exception of specialised mathematics processing (like say video
compression/conversion) where a sophisticated floating point unit can
easily outperform a lesser one at equivalent clock speed, that the
machine with the better memory bandwidth generally outperformed the
lesser.
Matt