Futuremark releases 3DMark03 patch - alleges NVIDIA cheats

[Ady]

could the FX owners please post detailed scores thanks. Not just overall. It would be nice to see what happens to individual game tests and the pixel shader test for everyone.

[chrismm28]

Originally posted by DaveM
HAHAHA, nVidia isn't cheating with my card.

5827 - before patch
5820 - after patch

Ok maybe I'm not gonna be so harsh since it didn't make a huge diff. on my card.
Looks like nvida are more clever than we thought !!! Its drivers have adaptive anti-cheat cheating mechanisms !!! ;) :D :angel2:

[digitalwanderer]

Originally posted by tertsi
I get now 4700 marks with 330 patch and 5800 Ultra (500/1000, Quality/Best image). without patch I get 5300 marks.

Well that's not THAT drastic of a hit, any other scores out there?

[solofly]

All this stress will shorten your lifes, folks. Over a useless time demo...

[curio]

Originally posted by solofly
All this stress will shorten your lifes, folks. Over a useless time demo...

Thanks for the advice Dr. Know

[DivotMaker]

Not sure how much this might have to do with this issue, but it appears that the new 3D Mark patch does not like the NV shader routines at all.

A representative at Nvidia questioned the validity of Futuremark's conclusions. "Since Nvidia is not part of the Futuremark beta program (a program which costs of hundreds of thousands of dollars to participate in), we do not get a chance to work with Futuremark on writing the shaders like we would with a real applications developer," the representative said. "We don't know what they did, but it looks like they have intentionally tried to create a scenario that makes our products look bad."

I am not making excuses for nVidia, but I sure did not expect the performance drop to be SO drastic...even 600 points overall is quite a big number....

[Ady]

Originally posted by BigBerthaEA
Not sure how much this might have to do with this issue, but it appears that the new 3D Mark patch does not like the NV shader routines at all.

A representative at Nvidia questioned the validity of Futuremark's conclusions. "Since Nvidia is not part of the Futuremark beta program (a program which costs of hundreds of thousands of dollars to participate in), we do not get a chance to work with Futuremark on writing the shaders like we would with a real applications developer," the representative said. "We don't know what they did, but it looks like they have intentionally tried to create a scenario that makes our products look bad."

I am not making excuses for nVidia, but I sure did not expect the performance drop to be SO drastic...even 600 points overall is quite a big number....


link?

[Clockwork]

What do you mean it doesn't like the nV shader routines??!??!

It's using standard DirectX9 shader/hardware calls. Just because nVidia can't toss in their custom (cheating) shader code as a replacement doesn't make Futuremark the bad guys..lol

[Kombatant]

Originally posted by Ady
link?

Seconded

[solofly]

Originally posted by Clockwork
doesn't make Futuremark the bad guys..lol

It sure doesn't but on the other hand also makes them USELESS in real world.

[Ady]

Possibly from Here (http://www.hothardware.com)

[DivotMaker]

Originally posted by Ady
link?

http://news.com.com/2100-1046_3-1009574.html?tag=fd_top

[Miester_V]

Just to speed things up...

linkage (http://news.com.com/2100-1046_3-1009574.html?tag=fd_top)

Edit: Damn, someone beat me to it...:mad: :)

[DivotMaker]

Originally posted by Clockwork
What do you mean it doesn't like the nV shader routines??!??!

It's using standard DirectX9 shader/hardware calls. Just because nVidia can't toss in their custom (cheating) shader code as a replacement doesn't make Futuremark the bad guys..lol

Shader "routines" might have been a poor choice of words. It could very well be that nVidia's drivers are not up to speed with respect to shader calls like ATI's is.

I never said Futuremark were the "bad guys".

[Ady]

Thats a huge drop at HH.

6067 -> 4886

edit: typo

[Ghola]

Originally posted by jAkUp
nonono... the reason we dont have ut2003 scores in our sigs is because anyone can fix that to give them any fps they want... walk around the map, turn off all bots, and stare at the floor.. gives you the most fps... to have an accurate way of testing various systems and video cards, we need a benchmark that is on a rail, that shows the benchmark the same way every time.

the problem is that if a benchmark is "on a rail" then video card companies can easily optimize... since they know exactly whats coming next... bah this is so confusing

Time to update your benchies huh?

[EMunEeE]

Originally posted by DaveM
Ok now it did drop

Grrrr nVidia.

Not as bad as those numbers.

5233 - just the gt4 is the main drop part, everything else sailed by the exact same.

DaveM

My Radeon 9700 Pro clocked at 351MHz GPU and 344MHz memory gets higher than a GeForce 5900 Ultra 256MB....thats gotta be sweeeett.:afro:

[hithere]

Originally posted by solofly
It sure doesn't but on the other hand also makes them USELESS in real world.

Yes, because in the real world, Nvidia replaces everyone's shader code with thier own and inserts clipping planes into games, right?

[Clockwork]

Originally posted by BigBerthaEA
Shader "routines" might have been a poor choice of words. It could very well be that nVidia's drivers are not up to speed with respect to shader calls like ATI's is.

I never said Futuremark were the "bad guys".

That's not Futuremarks problem considering they are following the DirectX 9 STANDARD

Oh, that's right. Don't you work for EA? They are technology partners with nVidia if I remember correctly. Some big press a bit back with their whole "The way it's Meant to be Played" campaign. Nope, you have no bias whatsoever....

:rolleyes:

[Ghola]

[QUOTE]Originally posted by DaveM
HAHAHA, nVidia isn't cheating with my card.

5827 - before patch
5820 - after patch

Ok maybe I'm not gonna be so harsh since it didn't make a huge diff. on my card. [/QUOTE


Whaa... they release another cheat already!!!? :p

[Ady]

another interesting point from the article here at hothardware (http://www.hothardware.com/hh_files/S&V/r9800256mb_gffx5900upd(2).shtml) is that 9800 scores go down on the standard test as expected 5834 -> 5747 yet on the AA+AF test it actually goes up. 2899 -> 2940 That does throw a little spin on the ATi issue. I guess we still have to wait and see what futuremark find.

[ChrisW]

So, nVidia did not pay any membership fees the entire time they were a beta partner? And Futuremark only charged dues when the benchmark was finished?:confused:

[A-KO]

I say **** Futuremark and their software and we use real games to tell the differences between the hardware. That way, if either company does place optimizations, at least they're to a usable benefit (higher FPS in a game).

edit by StealthHawk: do not circumvent the swear filter.

[Clockwork]

Originally posted by A-KO
I say **** Futuremark and their software and we use real games to tell the differences between the hardware. That way, if either company does place optimizations, at least they're to a usable benefit (higher FPS in a game).

Yes..once again it is Futuremark who are the bad guys...


:lame:

[solofly]

Originally posted by Clockwork
Yes..once again it is Futuremark who are the bad guys...


:lame:

No no... not bad guys but rather useless and alike.