How did NVIDIA cheat...exactly?

[Zane Krate]

Really, I'm very curious as to why a lot of people here are coming to the conclusion that NVIDIA cheated in 3DMark03. Where is the proof? How about referencing a link and pointing out the exact paragraph or sentence that proves beyond any doubt that NVIDIA cheated.

And why is it that people are automatically assuming NVIDIA's press release today clearly shows that they either bought out Futuremark or they threatened them with a lawsuit? Where's the proof?

[Sazar]

I am sure you can do a search through the topics in this forum... its all right there.. from about 2 odd weeks ago :)

[Zane Krate]

Originally posted by Sazar
I am sure you can do a search through the topics in this forum... its all right there.. from about 2 odd weeks ago :)

Not really, all I see if a bunch of dolts claming NVIDIA "clearly" cheated without showing a shred of evidence, and those who do actually take the time to post a link or source simply post the Extremetech article without actually explaining what it proves.

[jAkUp]

sheesh... this must of been posted 10 zillion times...

3DMark03 Patch 330 Released
Fri 2003-05-23

Saratoga, California USA – May 23rd, 2003

3DMark®03 incorporates many ground breaking new features and technologies and sets the standard for easily and objectively measuring and comparing modern PCs’ 3D graphics performance. Recently, there have been questions and some confusion regarding 3DMark®03 results obtained with certain NVIDIA drivers.

We have now established that NVIDIA’s Detonator FX drivers contain certain detection mechanisms that cause an artificially high score when using 3DMark®03. We have just published a patch 330 for 3DMark03 that defeats the detection mechanisms in the drivers and provides correct results.

Download the 3DMark03 Patch 330 here:

http://www.futuremark.com/download/...3patch330.shtml

A detailed audit report of the drivers and 3DMark03 is available at:

http://www.futuremark.com/companyin...udit_report.pdf

As with all Futuremark’s benchmarks, 3DMark®03 has been created in cooperation with the major PC industry players to provide users with the best possible and most reliable set of diagnostic tools.

Futuremark is regarded as the driving force that pushes the industry to raise the bar with respect to graphics performance and features. We are committed to deliver performance analysis in an accurate, fair and consistent manner. 3DMark®03 continues in that tradition.

Additional Information Please visit http://www.futuremark.com/products/3dmark03 for additional 3DMark®03 related information. 3DMark®03 is available for free public download at http://www.futuremark.com/download/?3dmark03.shtml

About Futuremark® Corporation
Futuremark® Corporation, formerly known as MadOnion.com™, is the leading provider of computer performance analysis software and services. Futuremark® is known around the world for its benchmark products, including the 3DMark® Series and PCMark™2002 (with more than 30 million copies distributed worldwide) and value-added services powered by a database of over 5 million real life benchmarking results. Futuremark® has offices in Saratoga, California and Helsinki, Finland. For more information, please visit http://www.futuremark.com.

© 2003 Futuremark® Corporation. 3DMark® and PCMark™ trademarks and logos, Futuremark® character names and distinctive likenesses, are the exclusive property of Futuremark Corporation. DirectX® is a registered trademark of Microsoft Corporation. All other trademarks are property of their respective companies.

Company Contact:

Futuremark Corporation
Phone: +1-408-517-0131
Mr. Tero Sarkkinen
Executive Vice President of Sales and Marketing
Email: tero@futuremark.com

[Sazar]

Originally posted by Zane Krate
Not really, all I see if a bunch of dolts claming NVIDIA "clearly" cheated without showing a shred of evidence, and those who do actually take the time to post a link or source simply post the Extremetech article without actually explaining what it proves.

if you read some of the first threads that contain pictorial evidence as well as futuremark's original audit pdf about this... before their 'arangement' with nvidia... it is clearly shown just what they were doing...

http://www.futuremark.com/download/?3dmark03.shtml

http://www.beyond3d.com/forum/viewtopic.php?t=5856

two threads/links that you may find interesting..

it will show 'clearly' that there IS evidence... at least to what we have to now refer to as 'optimizations'

[Zenikase]

Futuremark and Extremetech both wrote a good bit about exactly what nVidia was cheating in. The main problem was the fact that nVidia's drivers detected certain shaders in 3DMark03 and replaced them with their own in order to make it run better on their hardware. In addition, they altered the clip planes so that less scenery was rendered, thus producing higher scores.

[gordon151]

Originally posted by Zane Krate
Not really, all I see if a bunch of dolts claming NVIDIA "clearly" cheated without showing a shred of evidence, and those who do actually take the time to post a link or source simply post the Extremetech article without actually explaining what it proves.

Here's ur stinkin links u worm.

http://www.extremetech.com/article2/0,3973,1086025,00.asp

http://www.extremetech.com/article2/0,3973,1105259,00.asp

http://www.beyond3d.com/forum/viewtopic.php?t=5856

http://www.futuremark.com/companyinfo/3dmark03_audit_report.pdf

Edit: How did 3 posts appear in the like minute it took me to write this!! :(

[Zane Krate]

Originally posted by gordon151
Here's ur stinkin links u worm.

http://www.extremetech.com/article2/0,3973,1086025,00.asp

http://www.extremetech.com/article2/0,3973,1105259,00.asp

http://www.beyond3d.com/forum/viewtopic.php?t=5856

http://www.futuremark.com/companyinfo/3dmark03_audit_report.pdf

Edit: How did 3 posts appear in the like minute it took me to write this!! :(

LMAO! Are you trying to be one of those dolts by posting a link without actually explaining what it means? I've read all four of them. The first two links prove nothing, ET basically says that it appears NVIDIA is cheating but the evidence doesn't seem to clearly support that. ET also claims that ATI also wrote cheats for 3DM03. So in that case ATI AND NVIDIA are at fault.

Beyond3D is no different, they basically say something funny is going on but they can't put their finger on it (or maybe I missed something else they said in that very long thread?).

And of course FM has already recanted what they said earlier with their press release today, so I don't even know why you bother with that 4th link.

And why am I a "worm" for wanting to know how NVIDIA allegedly cheated?

[jAkUp]

and why are we dolts???

whatever that is..lol


the fm link is all you need. thats proof enough.

[Zane Krate]

Originally posted by jAkUp
and why are we dolts???

whatever that is..lol

LOL, not you in particular, just anyone who says NVIDIA cheated without taking at least a couple sentences to say exactly how.

For example, from what I've read so far, Zenikase's explaination looks the most possible, but just not 100% proven yet AFAICT.

Originally posted by jAkUp
the fm link is all you need. thats proof enough.

I read that, but since they recanted it how can I take it seriously?

[Sazar]

Originally posted by Zane Krate
LOL, not you in particular, just anyone who says NVIDIA cheated without taking at least a couple sentences to say exactly how.

For example, from what I've read so far, Zenikase's explaination looks the most possible, but just not 100% proven yet AFAICT.



I read that, but since they recanted it how can I take it seriously?

eh don't take it personal :)

most of us have discussed this as best as we can already... not really the energy to seek out all the links :D

I'll just post a few pics from b3d...

http://www.beyond3d.com/misc/3dmrendering/gt4_2.jpg

http://www.dragone.net/main/uploads/woodenbase.gif

just as an example of optimzations...

first picture is what happens when the camera goes off the rail... it is NOT supposed to happen and is a clear indication of something detecting the application and changing things around... (-edit- clipping panes)

second one is a basic showing of hackery in the shader tests...

[gordon151]

Originally posted by Zane Krate
LMAO! Are you trying to be one of those dolts by posting a link without actually explaining what it means? I've read all four of them. The first two links prove nothing, ET basically says that it appears NVIDIA is cheating but the evidence doesn't seem to clearly support that. ET also claims that ATI also wrote cheats for 3DM03. So in that case ATI AND NVIDIA are at fault.

Beyond3D is no different, they basically say something funny is going on but they can't put their finger on it (or maybe I missed something else they said in that very long thread?).

And of course FM has already recanted what they said earlier with their press release today, so I don't even know why you bother with that 4th link.

And why am I a "worm" for wanting to know how NVIDIA allegedly cheated?

I assumed too much of your understanding to assume that by simply reading the information offered in the links you would make inferences from the information offered. Also you're a worm because obviously you've been stuck under a rock for two years, otherwise you're too lazy to read the 10 to 20 threads on the subject which have around 100 people asking this very same question to only get it subsequently answered.

Oh well, whatever let someone explain it to you. I'm done.

[Clockwork]

Here's your proof:

http://www.futuremark.com/companyinfo/3dmark03_audit_report.pdf

I don't care if futuremark no longer calls it cheating. Any reasonable person can clearly see it is.

[StealthHawk]

There are so many threads on this, I will break it up into applicable sections for you to make for some organized reading:

New Cheats (44.03 & 43.51. Note the clipping planes in 3dmark03 GT2 actually appeared in 43.45, but we didn't know that until Futuremark's audit report)
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=11786
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=12354
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=12262
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=12024
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=11826
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=10056&pagenumber=2
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=11033

Old Cheats (43.45 and Older)
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=10128
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=8487
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=9206&pagenumber=3
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=9088
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=8614
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=7600
http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=9302

I think that sums it up nicely. If you have any questions about what you read, feel free to ask.

Other than that, it is fairly obvious that with the "old cheats" nvidia noticeably lowered the IQ. With the "new cheats" it is obvious they did a better job of hiding their cheating. At best there are some IQ sideeffects because of the clipping planes, but otherwise the IQ looks fine to the naked eye.

[Matthyahuw]

nVIDIA didn't cheat, they optimized...get over it...

[solofly]

Originally posted by Zane Krate
Where's the proof?

Non of us will ever see any proof unfortunately. Money is power and nVidia has a lot of it.

[StealthHawk]

Originally posted by Matthyahuw
nVIDIA didn't cheat, they optimized...get over it...

Cheap semantics. Here's something else Futuremark said today:Q: Does this mean what you called originally as "cheats" actually were acceptable "optimizations", and that you made a wrong decicion in releasing Patch 330 and the Audit Report?
A: By the definition of our benchmark and process, the optimizations are not acceptable. 3DMark scores are only comparable if drivers perform exactly the work 3DMark instructs them to do.

What is an invalid optimization? Isn't that a cheat?

[Hanners]

Originally posted by Zane Krate
I read that, but since they recanted it how can I take it seriously?

No, they haven't:

From Beyond3D (http://www.beyond3d.com/forum/viewtopic.php?t=6230)...

"First I must admit that there is very little I can comment about the joint statement between Futuremark and Nvidia due to legal aspects. What I can do is answer some frequently asked questions about this and quote some parts of the statement.

Please read the statement well and do not post hasty conclusion after reading only the first two paragraphs of the statement.

Q: Does this mean what you called originally as "cheats" actually were acceptable "optimizations", and that you made a wrong decicion in releasing Patch 330 and the Audit Report?
A: By the definition of our benchmark and process, the optimizations are not acceptable. 3DMark scores are only comparable if drivers perform exactly the work 3DMark instructs them to do.

The statement also says: Quote:
Because all modifications that change the workload in 3DMark03 are forbidden, we were obliged to update the product to eliminate the effect of optimizations identified in different drivers so that 3DMark03 continued to produce comparable results.


As earlier stated, we recommend using the latest build 330 of 3DMark03, with the 44.03 (or 43.51 WHQL) Nvidia drivers, or the Catalyst 3.4 ATI drivers. This way obtained 3DMark03 results are genuinely comparable as far as we know.

Q: What is the reasoning behind this statement?
A: Both companies want to end the public dispute that has been going on since we launched 3DMark03 in mid-February this year.

Q: Did NVIDIA pay you any money to make this statement?
A: No, they did not. Our companies had a mutual desire to end this dispute, and we are very pleased that we reached this agreement.

Q: Does this mean that in the future you will not make patches for 3DMark03 (or 3DMark2001) in order to reveal cheating?
A: We might release further patches to 3DMark03, if a need for preventing driver optimizations appear in the future.
_________________
Patric Ojala - 3DMark Producer"

[creedamd]

Originally posted by Zane Krate
Really, I'm very curious as to why a lot of people here are coming to the conclusion that NVIDIA cheated in 3DMark03. Where is the proof? How about referencing a link and pointing out the exact paragraph or sentence that proves beyond any doubt that NVIDIA cheated.

And why is it that people are automatically assuming NVIDIA's press release today clearly shows that they either bought out Futuremark or they threatened them with a lawsuit? Where's the proof?

How did they cheat? They made people like you think that the card was better than it was...

Proof?

You believed it and still do.

[R.Carter]

Originally posted by Zane Krate
Where is the proof? How about referencing a link and pointing out the exact paragraph or sentence that proves beyond any doubt that NVIDIA cheated.
If you can't understand from the audit report generated by FutureMark that Nvidia was taking shortcuts in their drivers for the benchmark, and that is cheating, well then I doubt anything can change your opinion.

Of course FutureMark wouldn't, and as far as I know, never did explicitly state that Nvidia was cheating since I suspect that would have gotten them into legal troubles.

Originally posted by Zane Krate
And why is it that people are automatically assuming NVIDIA's press release today clearly shows that they either bought out Futuremark or they threatened them with a lawsuit? Where's the proof? [/B]
Well, just where is Nvidia's press release I don't see anything in the Nvidia Press Room (http://www.nvidia.com/view.asp?PAGE=press_room) about this. Rather strange, considering that they have something dated June 3rd, 2003 up there. Maybe it'll be up tomorrow.

Again, if someone did provide "proof" I suspect that they would be violating the settlement agreement and that would be a "bad thing". Usually you have to sign a non-disclosure agreement (NDA) as part of the settlement. So FutureMark can't really say a whole lot about the incident so you will never get your "proof".

[Dazz]

Originally posted by Matthyahuw
nVIDIA didn't cheat, they optimized...get over it...
LMFAO ATi optimized and people including Futuremark considered it cheating does that change anything from what nVIDIA did? So ATi optimizes = cheat, nVIDIA 'optimizes' = no cheat, sorry you are contradicting yourself.

[Matthyahuw]

Originally posted by Dazz
LMFAO ATi optimized and people including Futuremark considered it cheating does that change anything from what nVIDIA did? So ATi optimizes = cheat, nVIDIA 'optimizes' = no cheat, sorry you are contradicting yourself. I didn't contradict myself...did you see me saying ATI cheated? Take your fanboy glasses off, what ATI did was fine by me as well, I'm all for getting better performance for nothing, no matter how it's done...

[hithere]

Originally posted by Matthyahuw
I didn't contradict myself...did you see me saying ATI cheated? Take your fanboy glasses off, what ATI did was fine by me as well, I'm all for getting better performance for nothing, no matter how it's done...

Dude, "better peformance for nothing" does not mean placing a map of where you know the camera is going to go in a benchmark into your drivers. They might as well have made you a movie about 3dMark and tried to pass that off as a "performance upgrade". If they played a DVD of Star Wars Episode 2 and said "this was all rendered in real time on an FX," and you believed them, would that mean that you had just gotten a massive performance boost, or would it simply mean that you are a gullible turkey?

[Megatron]

Originally posted by Matthyahuw
nVIDIA didn't cheat, they optimized...get over it...


You must have forgotten the "wink" or "smiley" emoticon on the end of that statement.

[Matthyahuw]

my motto is "whatever works", and I stick to it...

Can I get a show of hands here who base their video card purchase SOLELY off 3dmark?



None?



I rest my case...