ATI's Initial Responce to NVIDIA's Editors Day IQ Issues

Dave Baumann

Gamerscore Wh...
Moderator
Legend
We would like to respond to the recent allegations of benchmark cheating levied against ATI.

We take these allegations very seriously and are distressed by the implication that we are cheating our customers. At ATI we have taken a strong stance against application specific optimizations in benchmarks. Further, we never force optimizations that deliver lower image quality than the game developer has intended.

Three applications have been specifically questioned: UT2003, Halo and AquaMark3.

UT2003: We are working with Epic, to address a bug that has a slight impact on the image quality in the game and benchmark in certain situations. This is a known issue that has been recognized by both ATI and Epic as a bug. ATI is not cheating in any way in this application.

Halo: Although allegations of cheating have been made, no one has been able to find any example of decreased image quality in the game. In fact, many examples have been documented on the web where ATI cards and drivers are rendering a superior image to competing products. Again, ATI is not cheating in any way in this application.

AquaMark3: We are currently investigating our rendering in AquaMark3. We have identified that we are rendering an image that is slightly different than the reference rasterizer, but at this point in time we are unable to identify why that is. We believe that this does not have any impact on our performance. Our investigation will continue to identify the cause and resolve it as soon as possible. One point to note is that we render the same image using our latest driver (CATALYST 3.8) as we do with a driver that pre-dates the release of Aquamark3 by almost six months (CATALYST 3.2). Also, in all of our dealings with the developer of Aquamark3, at no point have they advised us that they are unsatisfied with the images that we are rendering. We do not have any application specific optimizations in our driver and we are not cheating in this application.

If you have further questions about the way RADEON graphics cards render in these application we encourage you to contact the game's developers directly.

We are disappointed that certain media outlets chose to perpetuate the allegations, made originally by a competitor, in their articles without an understanding of what was happening and without contacting us for comment.

We are committed to delivering the best gaming experience possible, without sacrificing image quality for increased performance in benchmarks.

I say initial in the title because I think there is more to come. Interesting that nobody decided to talk to ATI at all.
 
But to be honest toms did not say they were cheating, so what site actually did if any? Or is this just the generic over react statement?
 
fallguy said:
He did say, "certain media outlets". Probably pointing the finger at Toms.
You mean "She did say", I believe the statement was made by Patti Mikula. ;)

But you are right, I thought Dave was talking about other websites. :oops:
 
Okay then Lars - I presume you will at some point in the next 24 hrs or so browse this forum..... any reponse to why there was no attempt to verify or contact ATi?

I would be most interested to hear what you have to say on the apparent "situation" with regards ATi's, Epic's and the Aquamark developers positions on the matter.
 
Sxotty said:
But to be honest toms did not say they were cheating, so what site actually did if any? Or is this just the generic over react statement?

They did not have to. Alls they had to do is suggest. Then people ran wild with it. We had at least one thread posted here with the ATI cheats for the title. I saw simular threads on Rage, 3DGPU and NVNEWS. News got out, poeple ran wild with it. Had Tom's done a better job of confiming first then it would have been a non-issue...
 
digitalwanderer said:
Not true, I know for a fact that Hanners contacted them about it immedietly after reading about the accusations.

That's right - In fact, I still haven't heard anything back from them, not even so much as a copy of this statement. Oh well...

EDIT: I take that back... I did eventually get the same statement, after they got my e-mail address wrong the first time. :p
 
jb said:
Sxotty said:
But to be honest toms did not say they were cheating, so what site actually did if any? Or is this just the generic over react statement?

They did not have to. Alls they had to do is suggest. Then people ran wild with it. We had at least one thread posted here with the ATI cheats for the title. I saw simular threads on Rage, 3DGPU and NVNEWS. News got out, poeple ran wild with it. Had Tom's done a better job of confiming first then it would have been a non-issue...

Still people should not lie, lying and saying that Toms said "They are cheating" is worse than toms implying that they might be. People are diliberately misinforming people and that is wrong, so quit it please :).
 
Sxotty said:
Still people should not lie, lying and saying that Toms said "They are cheating" is worse than toms implying that they might be. People are diliberately misinforming people and that is wrong, so quit it please :).

Well, the phrase 'questionable optimisations' is not all that far away from saying cheating, there isn't really a huge distinction between the two in my mind beyond arguing the semantics.
 
I personally started hundreds of threads accusing ATi of being dirty cheaters after reading that Tom's article :D






j/k
 
Well I am touchy about news organizations taking things out of context, and the like. If you actually read what toms wrote, it really is not like they accused ATI of cheating they waffled around a lot and they implied it could be possible, but in the end they did not endorse any idea.

Even reading ATI's press release, they talk about accusations of cheating, but when it comes to specifics, they say perpetuate the allegations, not accuse. So even ATI realizes that sites did not accuse them, fanboys did. Sites perpetuated allegations, or in other words @ the worst they passed on what they were told w/o much comment. I did not see any large sites saying ATI is cheating, I only saw them say they might be and while I understand how this can tarnish ATI's image, I think it is really ridiculous to say it is some horrible blow, what it ended up doing is creating some publicity, and mayhaps a few folks will think that ATI did something bad b/c they will misremember, but more will think "gee whiz Nvidia is lame trying to besmirch ATI like that."

IMO of course :)/
 
Sxotty said:
Well I am touchy about news organizations taking things out of context, and the like. If you actually read what toms wrote, it really is not like they accused ATI of cheating they waffled around a lot and they implied it could be possible, but in the end they did not endorse any idea.

He insinuated it a lot though.

"This obviously saves memory bandwidth, leading to a higher framerate."

Does he have proof it does?

"Since the express purpose of this test is to test a card's memory bandwidth, an optimization such as this seems more than a bit questionable and smacks of cheating "

He says it is an optimization, but again, where is the proof? ATi said it is a bug. The same problem was in Cat 3.2', WELL before Aquamark3 came out.

"For now, we don't want to comment on the optimizations found in ATi's drivers."

Seems like he already did. By even calling them "optimizations" without knowing for sure, and insinuating that they gain performace with them. What is the average Joe going to think after what he said? I think its obvious.
 
Sxotty said:
But to be honest toms did not say they were cheating, so what site actually did if any? Or is this just the generic over react statement?

Tom's site perpetuated the allegations made by nVidia, imho.
 
Sxotty said:
But to be honest toms did not say they were cheating, so what site actually did if any? Or is this just the generic over react statement?
At best they helped perpetuate pure speculation. In sharp contrast to how they handled the proof when nVidia was found cheating. :rolleyes: To say that such comments would not tarnish an IHV's reputation is naive.
 
Look at the good with the bad

Well folks, at least the theory of cheating through optimization by these graphic card makers is a public idea. You think they'll screw around again? Yall are smart and stuff, the keen awareness to the situation by yourselves will keep a sizable balance I'm sure. Thanks.
Tyler
 
Now that's a quality press release: short and to the point. Assuming it's all true (and ATi were very specific to say they were not deliberately cheating), it reaffirms my view that ATi is just acting like a more trustworthy company lately. (Of course, as I've also previously said, ATi can afford to be honest, as they've had the best products for quite some time.)

Nice AM3 rebuttal, pointing out that pre-AM3 drivers render the same as post-AM3 ones. The fact that websites that repeated nV's accusations didn't also try this reinforces my position that, currently, consumer 3D card reviews just aren't professional.
 
Back
Top