ATI cheating

entium

Distinguished
May 23, 2004
961
0
18,980
There seems to be alot of news about ATI x800 cheating on trilear filtering. Are there any benchmarks where its turned on all the time compaired to the 6800 ultra?
 

GeneticWeapon

Splendid
Jan 13, 2003
5,795
0
25,780
There seems to be alot of news about ATI x800 cheating on trilear filtering.
<A HREF="http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=399365#399365" target="_new">http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=399365#399365</A>
<A HREF="http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=399611#399611" target="_new">http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&p=399611#399611</A>

<A HREF="http://rmitz.org/AYB3.swf" target="_new">All your base are belong to us.</A>
<A HREF="http://service.futuremark.com/compare?2k3=2216718" target="_new"><b>3DMark03</b></A>
 

splenda20

Distinguished
Mar 2, 2003
422
0
18,780
As far as I know, there aren't any benchies with Tri turned on all the time because ATI doesn't let you turn on or off their filtering technique.
So there is no way to turn on Tri 100% of the time, which is part of the controversy.
 
It can't be hacked in the registry for the FX anymore either, that was before people started posting those results in reviews.

There is no option to disable it because it is their trilinear method, not some alternate brilinear method to turn on and off like the nV method. What ATI could do is allow you the choice between the two algorithms, but that's different than enabling and disabling in the sense that your putting forth.

<b>Splenda</b>, if he bothered to read the threads GW posted then he'd have the answer to his questions as Digi, Wavey Dave, Hanners, and so many others looked into it very thoroughly, complete with pictures to help those mathematically challenged. :lol:


- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! - <font color=green>RED </font color=green> <font color=red> GREEN</font color=red> GA to SK :evil:
 

flamethrower205

Illustrious
Jun 26, 2001
13,105
0
40,780
Driver h4x4ge I like, maybe we should bring back the good ol' days of hacking NV's stuff...

SEX is like math. Add the bed, subtract the clothes, divide the legs, and hope you dont multiply
 

GeneticWeapon

Splendid
Jan 13, 2003
5,795
0
25,780
ATI doesn't let you turn on or off their filtering technique.
So there is no way to turn on Tri 100% of the time, which is part of the controversy.
Enabling colored mipmaps turns on true trilinear filtering you stupid moron. So I guess you could say that the answer that you gave this guy was wrong. Next time you want to complain that I didnt answer somebody....you'd better check yourself fool.


<A HREF="http://rmitz.org/AYB3.swf" target="_new">All your base are belong to us.</A>
<A HREF="http://service.futuremark.com/compare?2k3=2216718" target="_new"><b>3DMark03</b></A>
 

entium

Distinguished
May 23, 2004
961
0
18,980
hehe ATI just a dug a hole for itself. They were on the right track. Now anything they do is just a bigger hole.

Btw contacted ati won't do sh*t. They know if they turn on trilinear its going to kill them. They won't be able to match any of the 6800 numbers.

The reason they did this "optimization" was because at GDC the 6800 3dmark benches were released. At the time many people close to ATI said that the x800 won't be able to break the 12k mark.
 

AMD_Man

Splendid
Jul 3, 2001
7,376
2
25,780
I don't think ATI is cheating this time. Their answer to the accusations made sense. Trilinear filtering is about improving image quality. Who cares how you do it as long as you do it? If you've found a way to speed it up, then kudos to you! Sure, it's not pixel-per-pixel identical to full traditional trilinear but there is no rule that says everything rendered has to be mathematically identical between cards.

Intelligence is not merely the wealth of knowledge but the sum of perception, wisdom, and knowledge.
 

Xeon

Distinguished
Feb 21, 2004
1,304
0
19,280
But is it the same as the reference pixel, I think thats where the Nvidia guys finally caved on the idea that Nvidia was cheating or not.

With considerations that if a angle is at 45 degree's and the other is 36 degree's to the horizontal, the algorithim should see angle differences or distance differences and draw accordingly. It shouldnt need to be "adaptive" since the idea of the technology would already be dealing with different angles per frame.

Xeon

<font color=orange>Scratch Here To Reveal Prize</font color=orange>
 

Obtuse

Distinguished
May 21, 2004
377
0
18,780
It seems to me that a more efficient algorithm for trilinear filtering is perfectly ok. I mean, that's like criticizing a hybrid engine for not using "true" internal combustion, even though it gets you from point a to point b just as well as a normal one. Since trilinear filtering is just approximation of an image, who cares if they approximate efficiently? I have no pref for NVidia or ATI, but it sounds like we're criticizing good science here. I personally don't care if there's a f**king leprachaun inside my computer making the magic pictures appear on my screen, if it looks good, I'm happy. So, if there is definitive proof that IQ is degraded by this new algorithm, I'll be right there with you bitching at ATI. But right now, it seems like we're bitching over some kind of semantic misunderstanding.

"The only way to overcome temptation is to yield to it" - Oscar Wilde