Nvidia Cheats in 3Dmark2003? - Page 2

jmke · 20th May 2003, 10:09

Quote:

Why wasn't there more attention to detail in the Extreme Tech article? There could have been more screen shots or even map diagrams showing where the clipping planes and camera paths were. There could have been official statements from Furturemark. How did a copy of the Developer version of 3DMark 2003 get into Extreme Tech's hands? Certainly there would have to be a reason why the software was given to a Technology related site. I wasn't aware that Extreme Tech was developing hardware or software that would be benchmarked on 3DMark 2003. The developer license for 3DMark 2003 does cost quite a bit more than the Pro version of the benchmark. Somebody had to test for and find this anomaly before hand. Who's pulling the strings here? All I can do is look back in history and note that nVidia pointed out Quake 3 Arena wasn't being rendered properly on the Radeon 8500.

http://www.elitebastards.com/page.ph...d=1&comments=1

24th May 2003, 18:14

Update with reaction of futuremark (developper of 3Dmark) :
http://www.tweakers.net/nieuws/27117

Dutch site, links to THG (english)

->reaction of futuremark : new build in wich cheats don't work
->Ati performance drops with 2%, nVidia's performance drops 24,1 % ........

24th May 2003, 23:47

Oh I almost forgot to post this update from Futuremark, which apparantly prevents the use of optimized drivers; http://www.futuremark.com/download/?...patch330.shtml

27th May 2003, 13:47

and the story continues ..

http://www.xbitlabs.com/news/video/d...526040035.html

Now ATI admit that they did tweak the drivers too.

obviously not in the way nVidia did (score of Ati decrease with only 2%, that of nVidia more then 24%), but still..

I've read some comments, and they say that this isn't really cheating (cfr the benchmarks), but only doing some vertex shaders better.

->
< Gepost door hiostu dinsdag 27 mei 2003 - 08:56 - Score: 3
Ik vind zelf dat ATI niet gecheat heeft.

Wat ATI heeft gedaan zijn shader routines herkennen en ze in een betere vorm omschrijven. De uitkomst van de shader is precies hetzelfde alleen dan geoptimaliseerd voor de ATI hardware. Ze veranderen dus alleen de VOLGORDE van de shader instructies. Alles wordt met precies hetzelfde resultaat gerendered.

Daarnaast kan deze optimalisatie ook in games gebruikt worden. het gaat hier namelijk om veel gebruikte routines in shaders die ten allertijde herkent kunnen worden.

Dit is GEEN cheat.

Ook Tim Sweeney van Epic vind dat dit soort optimalisaties mogen en kunnen.

Kijk voor zijn reactie op http://www.beyond3d.com/#news6041
>

And one says that it isn't cheating at all, just a better type of drivers, and : that ATI never kept it secret that they improve DirX9.
c/p of comment added with the Catalyst 3.4 drivers :
DirectX 9 Performance Improvements
The following is a list of DirectX9 Performance Improvements noticed in the latest release of CATALYST™.

3DMark03 scores are improved as much as 8% across our entire DirectX 9 (dx9) product line
The difficult Mother Nature (dx9) test now runs as much as 20% faster than previous drivers

link : ATI Catalyst drivers 3.4

...

jmke · 3rd June 2003, 08:31

http://www.gamespot.com/pc/news/news_6029339.html
Nvidia didn't cheat?

Futuremark now has a deeper understanding of the situation and NVIDIA's optimization strategy. In the light of this, Futuremark now states that NVIDIA's driver design is an application specific optimization and not a cheat .

3rd June 2003, 10:07

hmm i think there's some cash involved

nvidia can't jeopardize there leaders position. they will have been (3d)marked as cheaters for sure.

3rd June 2003, 10:12

Thanks for posting this jmke, it's sign of intellgence to be open, and without accepting both sides of an argument, you go through life deprived of 50% of knowledge you otherwise would benefit from. In other words your only half as stupid if you at least listen to the other guy. (and I don't mean YOU personally)

I'm glad I posted that "patch" who knew it be the key which unravelled the mystery for Futurmark (I don't mean my posting it) I can under stand their change of heart considering Furturemark's benchmark is either based on current or "future" 3D code to be released. Which is why their often criticized as writing UNreal-world software. I beleive one reason behind theirnamsake change, was to enlighten gamers as to "Future" possibilities. Just like when DirectX 9 was released, people criticized IT, saying it wasn't yet necessary for current games (at that time). Point is this is a RAPIDLY evolving science/industry, hence the the constant game of "catch-up" played between hardware/software, and among companies.

OMG vegeta isn't that funny that was my first "instinct", but i didn't want to be contradictatory! Wow astute observation. I was unable to follow through on my initial thought, but breifly considrered Futuremark's possible dependence upon nVidia to have hardware with which to run their benchmarking software on. I thought, what if nVidia comes up with their own "3DMark" type benching software. That would surely damage Futuremark, and obviously nVidia has the capital to hold out longer. Not to mention a legal battle! Albeit a slightly different view from your own, it still comes down to money.

jmke · 3rd June 2003, 10:18

[H] Certainly is against Futuremark...

Quote:

Even based on FutureMark's own statements, none of the scores previous to their latest 3DMark03 Build 330 are to be relied on.
>>Can 3DMark03 be used as a reliable benchmark for DirectX 9 generation graphics cards?
Yes, with the new 3DMark03 build 330, it can.

Logically, that statement would seem to suggest to me that all scores previous to using build 330 are not reliable. We of course think that with or without the latest build 3DMark03 is not trustworthy. Do we wait for FutureMark to declare in writing that their next build is really the one that will give you reliable results. I would have to suggest that anyone using 3DMark03 scoring tests is irresponsible in one form or another.

that would be a great idea

the more competitors, the better the products

Quote:

I thought, what if nVidia comes up with their own "3DMark" type benching software. That would surely damage Futuremark, and obviously nVidia has the capital to hold out longer. Not to mention a legal battle!

3rd June 2003, 11:41

If I get right the summary of Futuremarks position for now is: "it's an application specific optimisation, not a cheat BUT Futuremark does not allow application specific optimisation".

That would still make it unfair to include it in a driver and it's just the same as before but reworded more to the liking of Nvidia (and it's lawyers/cashcows)

Personally I would've sticked to just calling it a cheat. They altered the way things looked which is simply not done.

If Nvidia just did it because they think 3dMark is not a fair benchmark they would've been better served by doing something about it publicly instead of cheating. What they did now was more something you'd do if you consider the benchmark fair but don't like your score.

BLMet

3rd June 2003, 15:33

That's true. If it's application specific, why not offer "Detonator UT3" or perhaps "Detonator Return to Castle Wolfenstien" for example. Why not begin including, "Driver optimized Game CD..." bundled with their videocards? Each may offer enhancements for a specific game, even cheats? Hell while their at it, why not write drivers which allow you to "roll forward" the score? Or just get the; You Win. Game Over your a Jedi Master! Oh I klnow why they don't do that, they'd have to create a division employing graphics designers for these drivers, and that would cost some serious skweepage. Funny though they did this where Futuremark is concerned, but for those os us who purchase their cards, nil. I don't know about you, but I can't imagine using any of the stupid cheats out now. In fact this comment may draw controversy, but I think copying files should be illegal, not promoted by the Kaaza's of the world. OK step off soap box now.

:grin:

27th May 2003, 13:47	#14
Dial_Up Posts: n/a	ATI too ? and the story continues .. http://www.xbitlabs.com/news/video/d...526040035.html Now ATI admit that they did tweak the drivers too. obviously not in the way nVidia did (score of Ati decrease with only 2%, that of nVidia more then 24%), but still.. I've read some comments, and they say that this isn't really cheating (cfr the benchmarks), but only doing some vertex shaders better. -> < Gepost door hiostu dinsdag 27 mei 2003 - 08:56 - Score: 3 Ik vind zelf dat ATI niet gecheat heeft. Wat ATI heeft gedaan zijn shader routines herkennen en ze in een betere vorm omschrijven. De uitkomst van de shader is precies hetzelfde alleen dan geoptimaliseerd voor de ATI hardware. Ze veranderen dus alleen de VOLGORDE van de shader instructies. Alles wordt met precies hetzelfde resultaat gerendered. Daarnaast kan deze optimalisatie ook in games gebruikt worden. het gaat hier namelijk om veel gebruikte routines in shaders die ten allertijde herkent kunnen worden. Dit is GEEN cheat. Ook Tim Sweeney van Epic vind dat dit soort optimalisaties mogen en kunnen. Kijk voor zijn reactie op http://www.beyond3d.com/#news6041 > And one says that it isn't cheating at all, just a better type of drivers, and : that ATI never kept it secret that they improve DirX9. c/p of comment added with the Catalyst 3.4 drivers : DirectX 9 Performance Improvements The following is a list of DirectX9 Performance Improvements noticed in the latest release of CATALYST™. 3DMark03 scores are improved as much as 8% across our entire DirectX 9 (dx9) product line The difficult Mother Nature (dx9) test now runs as much as 20% faster than previous drivers link : ATI Catalyst drivers 3.4 ...

3rd June 2003, 08:31	#15
jmke Madshrimp Join Date: May 2002 Location: 7090/Belgium Posts: 79,021	http://www.gamespot.com/pc/news/news_6029339.html Nvidia didn't cheat? Futuremark now has a deeper understanding of the situation and NVIDIA's optimization strategy. In the light of this, Futuremark now states that NVIDIA's driver design is an application specific optimization and not a cheat . __________________

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Driverheaven Nvidia Mobility Modder Released V1.1.0.0	jmke	WebNews	0	6th August 2009 13:12
NVIDIA Brings SLI Technology to Intel Bloomfield CPU Platforms	jmke	WebNews	1	14th July 2008 22:13
Nvidia cheats on 3DMark with 177.39 drivers	jmke	WebNews	1	23rd June 2008 23:36
Nvidia cheats in HD HQV test?	jmke	WebNews	0	31st August 2007 13:10
NVIDIA Brings the Power of SLI Technology to Intel Core 2 Duo Platforms	jmke	WebNews	0	6th June 2006 14:02
New NVIDIA nForce(TM)4 PCI Express Core-Logic Solutions Set Record Sales Growth	Sidney	WebNews	0	15th February 2005 18:11
NVIDIA Announces GeForce 6200 With TurboCache	jmke	WebNews	0	15th December 2004 15:57
NVIDIA Is the Only GPU Manufacturer	Sidney	WebNews	0	24th August 2004 05:44
NVIDIA Launches Revolutionary New Multi-GPU Technology	Sidney	WebNews	1	29th June 2004 07:21
Today NVIDIA introduced the Quadro FX 600 PCI	jmke	WebNews	0	19th March 2004 17:36

24th May 2003, 18:14	#12
Dial_Up Posts: n/a	Update with reaction of futuremark (developper of 3Dmark) : http://www.tweakers.net/nieuws/27117 Dutch site, links to THG (english) ->reaction of futuremark : new build in wich cheats don't work ->Ati performance drops with 2%, nVidia's performance drops 24,1 % ........

24th May 2003, 23:47	#13
Liquid3D Posts: n/a	Oh I almost forgot to post this update from Futuremark, which apparantly prevents the use of optimized drivers; http://www.futuremark.com/download/?...patch330.shtml

3rd June 2003, 10:07	#16
vegeta Posts: n/a	hmm i think there's some cash involved nvidia can't jeopardize there leaders position. they will have been (3d)marked as cheaters for sure.

3rd June 2003, 10:12	#17
Liquid3D Posts: n/a	Thanks for posting this jmke, it's sign of intellgence to be open, and without accepting both sides of an argument, you go through life deprived of 50% of knowledge you otherwise would benefit from. In other words your only half as stupid if you at least listen to the other guy. (and I don't mean YOU personally) I'm glad I posted that "patch" who knew it be the key which unravelled the mystery for Futurmark (I don't mean my posting it) I can under stand their change of heart considering Furturemark's benchmark is either based on current or "future" 3D code to be released. Which is why their often criticized as writing UNreal-world software. I beleive one reason behind theirnamsake change, was to enlighten gamers as to "Future" possibilities. Just like when DirectX 9 was released, people criticized IT, saying it wasn't yet necessary for current games (at that time). Point is this is a RAPIDLY evolving science/industry, hence the the constant game of "catch-up" played between hardware/software, and among companies. OMG vegeta isn't that funny that was my first "instinct", but i didn't want to be contradictatory! Wow astute observation. I was unable to follow through on my initial thought, but breifly considrered Futuremark's possible dependence upon nVidia to have hardware with which to run their benchmarking software on. I thought, what if nVidia comes up with their own "3DMark" type benching software. That would surely damage Futuremark, and obviously nVidia has the capital to hold out longer. Not to mention a legal battle! Albeit a slightly different view from your own, it still comes down to money.

3rd June 2003, 11:41	#19
BLMet Posts: n/a	If I get right the summary of Futuremarks position for now is: "it's an application specific optimisation, not a cheat BUT Futuremark does not allow application specific optimisation". That would still make it unfair to include it in a driver and it's just the same as before but reworded more to the liking of Nvidia (and it's lawyers/cashcows) Personally I would've sticked to just calling it a cheat. They altered the way things looked which is simply not done. If Nvidia just did it because they think 3dMark is not a fair benchmark they would've been better served by doing something about it publicly instead of cheating. What they did now was more something you'd do if you consider the benchmark fair but don't like your score. BLMet

3rd June 2003, 15:33	#20
Liquid3D Posts: n/a	That's true. If it's application specific, why not offer "Detonator UT3" or perhaps "Detonator Return to Castle Wolfenstien" for example. Why not begin including, "*Driver optimized Game CD..." bundled with their videocards? Each may offer enhancements for a specific game, even cheats? Hell while their at it, why not write drivers which allow you to "roll forward" the score? Or just get the; You Win. Game Over your a Jedi Master!* Oh I klnow why they don't do that, they'd have to create a division employing graphics designers for these drivers, and that would cost some serious skweepage. Funny though they did this where Futuremark is concerned, but for those os us who purchase their cards, nil. I don't know about you, but I can't imagine using any of the stupid cheats out now. In fact this comment may draw controversy, but I think copying files should be illegal, not promoted by the Kaaza's of the world. OK step off soap box now. :grin: