ati cheating in crysis benchmark?

[saaya]

that would explain why a 5870 beats a 480 in crysis...

german article:
http://www.pcgameshardware.de/aid,74...e/Test/?page=3

google translations from german to english are a nightmare... the pics speak for themselves, despite setting anisotropic 16x filtering in catalyst control panel or in the game itself, atis latest catalyst drivers actually force a mix of bi linear and tri linear texture filtering, resulting in a few percent performance boost, but flickering and blurred textures. welcome back PRE quake3 texture filtering! its been a while!



i hope ati can fix this "bug" asap...
pretty lame... as if those few extra percentage points would make a difference anyhow... instead they force optimizations you cant turn off in any way... talk about bending over your customers... :/

things are looking so good for ati lately... i cant believe they did something like this... they dont need this... really lame...

i wonder if the same texture filtering "bug" is happening in other games... so ironic if you think of how anandtech and other reviewers were tricked into thinking atis texture filtering is better because its truly angle independant... well, it CAN be... unless ati thinks they need a few more fps here and there... i thought we were beyond those kind of "tweaks" :/

[Luka_Aveiro]

This can't be true because we are not talking about nVidia here.

[Helmore]

I hope they fix that quickly, I might forgive them if they do.

[XSAlliN]

ATi!? No, that's almost impossible, since they are the good guys.

[randomizer]

This can't be true because we are not talking about nVidia here.

Agreed. Some NVIDIA marketing guy got paid to make up this rubbish.

[saaya]

if it turns out ati did this in several games and continues to tweak reducing image quality to get closer to gf100 performance then id rather buy an nvidia card, even if its more expensive and hotter...
h3ll, thats why i stopped buying nvidia cards a few years ago, cause they kept lowering iq to get a few more fps in benchmarks (geofrce5+6 days)

if ati beats nvidia in crysis by 1% or loses by 2%, who cares?
as if that actually would impact people enough to buy one card over the other...

[saaya]

No way. Only Nvidia can make this stuff up. I bet saaya was paid handsomely also.

no, ati didnt send me my last cheque so i jumped ship

seriously tho, whats up with the comments?
you make it sound as if xs was ati biased... why would it be? some mystical reason? or does all ati staff post here? or does ati hire 50 forum trolls to spread ati pr?
come on... nvidia was so clueless it couldnt find its own schlong in the dark for the past months, the nvidia bashing was for a reason, not out of bias...

back on topic... *looks at ati* LAAAMMMEEEE!

[grimREEFER]

well, this would explain the curious discrepancy between far cry 2 and crysis benches

[Helmore]

well, this would explain the curious discrepancy between far cry 2 and crysis benches

I don't think you can make any comparison between the Far Cry 2 and Cysis benchmarks. Those 2 games use completely different engines.

[saaya]

well, this would explain the curious discrepancy between far cry 2 and crysis benches

well, they are based on different engines...
but it would def explain why ati cards rock in crysis and fall back up to 30% in other games.

----------------------------GTX285 and GTX480 ---------------------------------------------------------------- 5870-----------------------------------


difference:

[annihilat0r]

In the end, according to Google Translate it says "in view of the recent perf. increases by Catalyst drivers, these should not go unmentioned"

So this doesn't happen in prior Catalyst drivers?

[Sadasius]

We need a 3rd company to make decent video cards. Would help keep others in line and keep the out. Bah who am I kidding. It would just add to the pile of wouldn't it?

Well ATI's turn....

[Hornet331]

easy to prove, does it also happen with other drivers? 9.12, 10.1 or any other driver then 10.3a?

If not -> bug, kthxbye next please.

If true -> shame on ati.

[annihilat0r]

daamn I just installed 10.3

[saaya]

easy to prove, does it also happen with other drivers? 9.12, 10.1 or any other driver then 10.3a?

If not -> bug, kthxbye next please.

If true -> shame on ati.

even if it only happens on 10.3a, how can you call it a bug?
its a very convenient bug isnt it?
that just so happens to show up in the latest driver reviews had available before the nda for the 400 series ended... and it just so happens that this bug boosts their performance to a psychological win over the 400 series in a popular benchmark... come on...

i HOPE its only present in 10.3a... but im pretty sure its also there in 10.3... and ill be really pssed if its still there in 10.4

[XS2K]

Seeing that Fermi didn't live up to expectations Jenny from the block aka. Jen Hsung Huang forced Crytek (since they are a member of the evil club TWIMTBP) to degrade IQ for ATi.

[DOM.]

well, they are based on different engines...
but it would def explain why ati cards rock in crysis and fall back up to 30% in other games.

----------------------------GTX285 and GTX480 ---------------------------------------------------------------- 5870-----------------------------------


difference:

to me the 5870 looks better

[Rattle]

crysis

is that a benchmark ?

[naokaji]

We need a 3rd company to make decent video cards. Would help keep others in line and keep the out. Bah who am I kidding. It would just add to the pile of wouldn't it?

Companies exist to make money and to do so they will do pretty much anything to make their product look good, in other words, a 3rd company would indeed just add to the pile of bs.

Anyway, does anyone actually play crysis regularly or is it just used as a benchmark like 3dmark?

[Xcel]

If the posted picture comparison is the best you got I'd rather take the extra fps since I don't see any difference. Unnoticeable optimisations

[Frontl1ne]

Have there been any comparison screenshots between pre 10.3 drivers and 10.3 drivers? The article is taking a long time to load since I'm capped until 2AM.

If this is what ATi's been doing all along, I don't see what the particular problem is since no one's ever really complained about ATi's AF quality in Crysis...

[Xello]

you make it sound as if xs was ati biased...

Woah really what a shocker

[Oliverda]

PCGH? Curiously they forgot to put the HD 5970 into theirs Fermi review...

[annihilat0r]

PCGH? Curiously they forgot to put the HD 5970 into theirs Fermi review...

yeah... they must've totally faked that screenshot. In fact, I believe they first fabricated the "banding difference" pic with the black background and dots, and then applied it on the original screenshots so it would look like banding. cmon guys, we are talking about ATI here. what'd they ever to do us ?

[kemo]

So they found an AF Bug and still benched the game with that Bug , it is pointless to add that bench with that bug available , should have rolled the drivers back till it is fixed or at least mention it in the graphs ..