|
AGP
Aperture performance Comparison
As
with all the other benchmarks the results displayed
were the middle of three runs with any that deviated
sifnificantly rerun. The PC was rebooted each time,
which is obvious really otherwise I couldn't have
changed the aperture size for each run :)
|
256MB AGP Aperture
|
|
Ti4200P-Turbo
|
 |
|
Ti4680P-Turbo
|
 |
|
128MB AGP Aperture
|
|
Ti4200P-Turbo
|
 |
|
Ti4680P-Turbo
|
 |
|
64MB AGP Aperture
|
|
Ti4200P-Turbo
|
 |
|
Ti4680P-Turbo
|
 |
|
32MB AGP Aperture
|
|
Ti4200P-Turbo
|
 |
|
Ti4680P-Turbo
|
 |
|
16MB AGP Aperture
|
|
Ti4200P-Turbo
|
 |
|
Ti4680P-Turbo
|
 |
Looks
like nothing has changed. Above 32MB there's nothing
between the results as the AGP aperure size increases.
It may be worth experimenting with your own system
to find the optimal setting if it really matters that
much to you but the aim here was to see if AGP8X cards
coupled with AGP8X motherboards were in some way overly
sensitive to AGP aperture setting, and the conclusion
it seems is that they're not.
While
we're in experimental mode, you may remember a couple
of years back I wrote an article in which I looked
at the principles behind Fast Writes in video cards
and then, after running a few benchmarks came to the
conclusion that enabling fast writes was actually
reducing performance rather than increasing it? Well
guess what, nothing's changed there either! I'll demonstrate
the effects of enabling Fast Writes on the next page
after having switched to the KT333 Albatron KX400
Pro motherboard but I can assure you exactly the same
results were observed on this KT400 motherboard too.
|
