Futuremark releases 3dmark 2003 patch. nVidia officially cheated.

[PreservedSwine]

Found this over @ NVnews...

http://www.futuremark.com/news/?newsarticle=200305/2003052308#200305/2003052308



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/?3dmark03patch330.shtml

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

http://www.futuremark.com/companyinfo/3dmark03_audit_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

EDIT: Corrected Links

 

[SafirXP]

So has anyone tested it yet? Would love to see a significant drop in scores after installation of this patch. Wonder how NVidia will react.

FutureMark too should consider making some changes with their next graphics/gaming benchmark. With some rethinking & user suggestions, a better benchmark program can be made, I believe.

 

[james.miller]

Ahhhhaaahaaa.Ahhahaahaha.Ha

gotta laught at that. Not un-expected though. Now i want to see those new scores. Question is, are people actually going to use the patch if they have nvidia cards?

 

[Evnas]

Question is, are people actually going to use the patch if they have nvidia cards?

Probably not, but im sure that on the ORB it will tell you what version the person used

 

[IR1]

Wonder if they can ban you from posting scores until you do get the patch. Kinda force the patch on you.

 

[baraka]

It looks like the drop in NV's 3dmark score will be about 25%!

http://www.extremetech.com/article2/0,3973,1103987,00.asp

 

[ninthebin]

maybe future releases should make Nvidia cards perform worse...so then when Nvidia cheats it will cancel everything out

 

[james.miller]

ouch

 

[quegyboe]

Harsh

 

[Pinky]

lol

yeah, that will sting... but maybe it will learn 'em

 

[glock19owner]

So I guess my 1,339 score on a 8RDA+ with a GF3 Ti200 and a 1700@2350 using the 44.03 (WHQL cert.) is not a real score...I find that hard to believe...

And didn't ATI do something like this a few years ago also...never heard of a patch to solve that issue...

 

[DaddyB]

ATi did it with quake 3 when the 8500 was released and not to defend it but they did fix their drivers and the new drivers ran quake faster then the 'cheating' drivers. They more or less did it hoping not to get caught and knowing that they would get rid of the cheat and have better performance when they optimized the drivers some more, they knew their card could get they level of performance that the cheat gave them once they fixed their drivers up a bit. I mean when the 8500 was up against the GF3 it lost but now the 8500 is a faster card thanks to some driver revisions, even Nvidia cards get better with more mature drivers.

Nvidia did it because without the cheat their card doesn't compete with the ATi cards although they will claim that they were showing how 3dmark is a useless benchmark no doubt.

 

[kamz]

so this now throws the 5900 ultra into some serious doubt... what is now the real performance of this card against the 9800pro will very interesting indeed...

 

[PreservedSwine]

And didn't ATI do something like this a few years ago also...never heard of a patch to solve that issue...

No, they didn't.
I think you're reffering to Quake3?
Ati was using application detection, but the *optimization* had been in no less than 7 previeous driver prior to the release of the 8500. The optimization did not cause any issues on the ATI64DDRVIVO, the problem was that the exsisting optimization caused an issue on the just released R8500 (lower quality IQ) wich ATi did not apparently notice. The following driver release from ATI fixed the IQ problem, and *increased* Quake3 performance. I wouldn't call it a cheat.....

What Nvdia has done is particularly scandalous. The severley limited the workload of the GPU in 3dM03 by using clip planes...They discarded shader routines, and had their driver run shader routines that aren't even in the benchmark!

What Are The Identified Cheats?


Futuremark’s audit revealed cheats in NVIDIA Detonator FX 44.03 and 43.51 WHQL drivers. Earlier GeForceFX drivers include only some of the cheats listed below.

1. The loading screen of the 3DMark03 test is detected by the driver. This is used by the driver to disregard the back buffer clear command that 3DMark03 gives. This incorrectly reduces the
workload. However, if the loading screen is rendered in a different manner, the driver seems to fail to detect 3DMark03, and performs the back buffer clear command as instructed.

2. A vertex shader used in game test 2 (P_Pointsprite.vsh) is detected by the driver. In this case the driver uses instructions contained in the driver to determine when to obey the back buffer
clear command and when not to. If the back buffer would not be cleared at all in game test 2, the stars in the view of outer space in some cameras would appear smeared as have been reported in the articles mentioned earlier. Back buffer clearing is turned off and on again so that the back buffer is cleared only when the default benchmark cameras show outer space. In free camera mode one can keep the camera outside the spaceship through the entire test, and see how the sky smearing is turned on and off.

3. A vertex shader used in game test 4 (M_HDRsky.vsh) is detected. In this case the driver adds two static clipping planes to reduce the workload. The clipping planes are placed so that the
sky is cut out just beyond what is visible in the default camera angles. Again, using the free camera one can look at the sky to see it abruptly cut off. Screenshot of this view was also
reported in the ExtremeTech and Beyond3D articles. This cheat was introduced in the 43.51 drivers as far as we know.

4. In game test 4, the water pixel shader (M_Water.psh) is detected. The driver uses this detection to artificially achieve a large performance boost - more than doubling the early
frame rate on some systems. In our inspection we noticed a difference in the rendering when compared either to the DirectX reference rasterizer or to those of other hardware. It appears
the water shader is being totally discarded and replaced with an alternative more efficient shader implemented in the drivers themselves. The drivers produce a similar looking rendering, but not an identical one.

5. In game test 4 there is detection of a pixel shader (m_HDRSky.psh). Again it appears the shader is being totally discarded and replaced with an alternative more efficient shader in a similar fashion to the water pixel shader above. The rendering looks similar, but it is not identical.

6. A vertex shader (G_MetalCubeLit.vsh) is detected in game test 1. Preventing this detection proved to reduce the frame rate with these drivers, but we have not yet determined the cause.

7. A vertex shader in game test 3 (G_PaintBaked.vsh) is detected, and preventing this detection drops the scores with these drivers. This cheat causes the back buffer clearing to be
disregarded; we are not yet aware of any other cheats.

8. The vertex and pixel shaders used in the 3DMark03 feature tests are also detected by the driver. When we prevented this detection, the performance dropped by more than a factor of
two in the 2.0 pixel shader test.

 

[IR1]

So I guess my 1,339 score on a 8RDA+ with a GF3 Ti200 and a 1700@2350 using the 44.03 (WHQL cert.) is not a real score...I find that hard to believe...

Well Im pretty sure the cheats in the 44.03's only cheated for the Fx's...

but... on the off chance it also cheated for the Geforce4's and 3's. You better believe your score isnt real, because Nvidia was trying to fool you the consumer into buying a product by making it seem faster than it really is. This is fraud, almost criminal imo.

Now the quake3 8500 incident, while it isnt nearly as blatant or insidious, is no justification for Nvidia to go do something as stupid as this.

Its Nvidia's turn to clean up their act. Since the quake 3 incident ATI has been on the ball with every card.
Frankly Nvidia is a mess, they still have more market share, but there business partners arnt to happy to say the least.
I dont know if I can ever trust em again.

 

[ktn]

Ahh...slowly...losing....nvidia...fan...boy....status....ArGhhH!

 

[IR1]

Just try and stay away from being a fanboy period!

I try not to get to atached to ATI. Though they been so great lately I have a hard time not sticking up for their hardware. My last card was a gf3 ti500 and I loved it. Its in my bros computer now.

 

[PhobMX]

so this now throws the 5900 ultra into some serious doubt... what is now the real performance of this card against the 9800pro will very interesting indeed...

ditto... watch all the nvidiots look at their scores fall back and cry LOL

 

[Cowboy X]

LOL .............. I hope this puts to rest all of those who insisted that it was a driver bug . Or that we should all wait and see , well the truth is here and crow is being served .

 

[Fever]

Wow... Very pathetic of nVidia if you ask me. I have a feeling I won't have an nVidia card in my rig anytime soon.