More on ATi 'optimisation' - Further AquaMark 3 analysis
Jaybo
I'd like to know what proof he is trying to find, that could change the outlook of the article quite a bit. Is it to defend or blame? Only think I wonder about is if color palette has been taken into consideration instead of just gamma. Can we assume that the rest of the tests outside of Aquamark will be further investigated?
thomas997
"ATi appears to have some kind of driver bug (possibly affecting gamma levels) that makes things appear darker than they should. There seems to be no performance benefit from doing this, so to call it a 'questionable optimisation' seems wholly inappropriate in light of this" Thats what I was thinking before, how the f*ck does brightness affect the work that has to be done?
digitalwanderer
Some people were claiming that ATi's cards weren't rendering the same number of pixels; that it wasn't just darker it was not doing the right amount of work. Hanners did a great job of proving 'em wrong and showing that it is indeed just a difference in gamma. :)
Jaybo
Apparently not, no explanation to any further tests so far. My my, don't we love to jump to conclusions to take a stab at a popular target...
neo-n
trouble is your looking for things that just arent there, all ATI 'cheats' that have been proved are so minor its ridiculous. Most are bugs or just quirks of the hardware anyway.
neo-n
Nvidia are clutching at straws, i would expect then to turn their attention on XGI soon, they will probably be an easier target.
Jaybo
much like how optimisations aren't cheats?
neo-n
No some optimisations are fine, others are cheats. Cheats gain useful speed and reduce image quality. Nvidia has lots, ATI has few. I dont call something that gives 1 or 2 FPS a cheat, but something that gives 30 or 40 is. Of course its all a matter od percentage, so less than 5% isnt, but 10% or more is. Everyone has different standards, but those are mine.
