Nvidia Busted-Cheating With Their New FX Drivers

[Nemesis77]

Originally posted by: Vegetto
meh, i personally dont care if one graphics company or another 'cheats' their video cards or drivers just to get a better score on 3dmark. its just a general video card benchmarking program. all that matters is how the cards perform in real games

well.. thats my opinion

You do realize that if NV can do this with 3DMark, they can do the same with any other timedemo? All the "real-life" benchmarks run a demo where the camera is on a "fixed track". Since NV can do this on 3DMark, they can do the same on UT2K3 Flyby, Max Payne "Shooting Alex" etc. etc.

 

[Nemesis77]

Originally posted by: Shamrock
Now the dilemma...Did NVidia cheat in 3dMark 2k3, or are they showing us "flaws" in the program? I would see it as it's a flaw in 3dmark...because AMD and NVidia both have been outspoken about 3dMark.

NVidia isnt cheating at anything, they are exposing a flaw in the program. At least NVidia didnt "cheat" in a game result, and only exposed a flaw in a SYNTHETIC benchmark :\

If they are doing that, why don't they just say "We are doing this so show how flawed 3DMark is". Instead they said "Errr.... It's a bug in the drivers. Yeah, that't is, a bug". And besides, how does 3DMark suck, since it obviously has the needed tools (movable camera) to expose this kind of cheating?

 

[Nemesis77]

Originally posted by: Shamrock
NVidia isnt cheating, they simply exposed a flaw in 3dMark2k3. They knew tech sites would catch on.

Oh I see. When Ati cheats, they cheat. When NV cheats, they are just "exposing the flaws in the benchmark". Yeah, right! If they are doing that, why don't they just say it? Instead they claimed that it was a bug in the drivers.

NV's official explanation:

nVidia believes that the GeForceFX 5900 Ultra is trying to do intelligent culling and clipping to reduce its rendering workload, but that the code may be performing some incorrect operations.

Funny, I fail to see any mention of "Yeah, we did this on purpose to show how much 3DMark sucks"

 

[Rand]

Originally posted by: jliechty
Well, I guess (if that's true) that the nvidiots won't be able to point a finger at that fanATIcs anymore now that both companies have cheated at one time or another in some benchmark or application.

I always found the idea that one company is supposedly particularly moral and upstanding and would never cheat as being rather amusing.
Both ATi, and nVidia have publically admitted to cheating in the past as far back as the old WinBench98 fiasco shortly after Intel started the mess with the I740.

If they've done it once and admitted to it, I for one wouldnt be at all surprised if they've both done so numerous other times and gotten away with it.

 

[WicKeD]

Honestly, I dont care. Extremetech is one of the worst review sites out there. I am not saying they are wrong, but I do not believe anything they write.

 

[Live]

Originally posted by: Shamrock
NVidia isnt cheating, they simply exposed a flaw in 3dMark2k3. They knew tech sites would catch on.

and who's to say the"gamegauge" isnt ATI optimized? hmm... Extremetech seems biased towards ATI on the FIRST page

before you start calling me a fanboy or an NVidiot, I have owned every major video card, EXCEPT Matrox, I've even owned a Rendition based Diamond V2200....so I am completely neutral.

What exactly is this "flaw" you are talking about? They have not shown us any flaw. They have given us the consumers an official set of drivers that manipulates 3DMark03 in a way that its not intended to do, and in no way did they tell us about it. This gives them higher scores in one of the most commonly used benchmarking programs around. That is called cheating and lying not exposing a flaw.

Your "theory" doesn?t hold up at all. As you know Nvidia has openly spoken out about 3DMark03 before. If they would like to expose a flaw they could have and should have just published their findings. They have done it before, what stopped them this time? Knowingly not rendering stuff and using the memory on the card in a way that would not work in any other benchmark or game, and not telling about it but instead claiming 30% better performance from the drivers is exposing a flaw?

Nvidia is a big corporation and I don?t believe for a second they would expose a flaw in a benchmarking tool by cheating it with there latest sets official drivers, in a hope that someone would find out and then publicise there results and thank Nvidia for finding the flaw. Since it?s a sure way to loose money why in the world should they? It?s much more credible that Nvidia did not know of the latest implementation of a free camera mode in the developer version 3DMark03 since they are not part of the futuremark team. There is no way of finding these cheats without that free camera mode that has not existed before in any 3DMark version.

We as a community should not stand for these big corporations trying to lure our hard earned cash from us, with lies. Some here seems to think that if lying is good for business its alright. Maybe for Nvidia stock owners, but not for us Nvidia products owners. Competition is good for all of us. Unjust competition is bad for all of us as it leads to higher prices and lower quality. Enron did try that lying road and it didn?t go to well for them or for the owners, employees or the tax payers for that matter. (no other comparison). Nvidia have acted without any respect to us, the consumers and users of Nvidia products. I say throw them out of the temple the bastards.

And by all means call me a fanboy. Except from the great Voodoo cards from 3DFX I have only owned Nvidia cards. I have liked Nvidia for many years. That does not mean I will look less hard on a thing like this. On the contrary I have given them my money for a product and the promise of serving me with excellent drivers for it. This is not what they have done.

 

[Schadenfroh]

we cant trust ati for cheating at quake3, we cant trust nvidia for they cheated at 3dmark2003. the only company we can trust is... well.. TRIDENT!, yeh they have never manipulated benchmarks.

 

[RobsTV]

Doesn't make sense.
Nvidia is able to incorporate into their drivers enough info to hardcode frame info for all of 3dm2k3 into those drivers????
Yet these have no adverse affect on other games not "on a rail", and keep size small?
Yeah, right.
If this is true, then they surely have the best driver techs around, and that is a good thing!
Even if you throw out the beating the Nvidia card did to Ati in that one test, in all the other tests the cards were fairly equal.
Point is, even if equal, as other tests show, their is nothing wrong with that. Add to that the magic driver techs, and looks good.
It's not like they are trying to say that all results were from cheating, just one test.
And even that is questionable.

But, I still think the FX sucks in comparison to previous Nvidia releases.

 

[Nemesis77]

Originally posted by: RobsTV
Doesn't make sense.
Nvidia is able to incorporate into their drivers enough info to hardcode frame info for all of 3dm2k3 into those drivers????
Yet these have no adverse affect on other games not "on a rail", and keep size small?
Yeah, right.

Well, there are people who think this is blatant cheating, and they know ALOT more about 3D than you do...

 

[Mem]

Oh I see. When Ati cheats, they cheat. When NV cheats, they are just "exposing the flaws in the benchmark". Yeah, right! If they are doing that, why don't they just say it? Instead they claimed that it was a bug in the drivers.

To be honest I don`t give a damn about 3dmark benchmarks,I don`t know about you but I use my drivers for real games etc,infact as long as my games run great I`m one happy gamer,3dmark is not even on my top 1000 game list .

It always amazes me when people say "I got 20 points this or 100 points that more with these drivers" who cares ,use the card and drivers for real games that`s why people buy it for.

As for cheating well I don`t know if it`s true or not ,but don`t care,it`s pretty hard to care about a piece of software that`s not important in my books and hardly used.

Lol..I only use 3dmark to check my 60Hz refresh rate fix is working in WinXP when I install new drivers.
Bottom line is I want drivers to work fine in real games,that`s all I care about.

 

[joe4324]

I'll withhold my opinoin untill more time has passed and this issue has been looked at even closer. BUT I will say this. I have a 9500PRO in my system right now. Its a great card. BUT I really dont give a rats @ss about 3dmark even if it was a totally intentional move by nvidia to deceitfully inflat the scores, I would still wouldnt even put it on the same page or even in the same chapter as the qauck3 disaster.

That little "bug" ati got caught with was both, Harmful to one of the most popular real-world games as it reduced its image quality to inflate scores and was BLATANTLY hand coded in.

ahhh hell, what do I care. I really dont trust any for-profit corporation anyway. who ever scores the best in real world games for the least amount of money is what i'll buy reguardless of these issues. Like I'm running an ATI card right now. even after Qauck3 issue...

 

[nRollo]

Oh NO! They CHEATED at 3dmark2003!!!!!
<faints dead away>

LOL who gives a fat rats a$$? I don't play much 3dmark2003, if it's faster on games with equivalent IQ, that is all that matters to me.

As far as the fallacy of my "argument" (more a "comment" about the the quality of the site) I was just pointing out they have a strange way of doing business.

I assume they're an offshoot of CGW (rag) with Salvator and Gameguage? I used to subscribe to CGW, when it was more than 15 pages an issue and had Lloyd Case. They were pretty sad with the "Look it runs this bench at 11fps, not 8 like it's predecessor" BS. Even if you're going to say "we want to max out these cards" I think you need to run them at PLAYABLE levels to give the reader some idea what to expect. What can I draw from their review? That neither of these cards run games acceptably at these resolutions/AA/AF settings? That's nice. Also about as useful as running the benches at 6X4X256colors, no AA/AF.
Where is their performance at settings that can offer smooth gameplay? That would be more useful to me.

 

[ginfest]

No matter what the "truth" is in this case, one thing stands out since the 5900U/9800 anouncements and benchies came out-
All of a sudden, because the 5900U doesn't appear to be the failure that the 5800 was, and actually beats the ATI offering in some cases, there is suddenly a rash of crap- " better offering from ATI taping out", "sites and reviewers being accused of dishonesty etc", "can't buy 5900U yet but 9800 Pro is available"and now "NV driver cheats".
What, ATI wins 1 round in the 3D vc wars and NV can never make a "better" card?
Why don't we wait for final silicon/drivers and actual DX9 games to be released b/4 worrying about who's best?

Mike G

 

[Nemesis77]

TO ALL OF YOU WHO SAY "IT'S JUST 3DMARK, WHO CARES?":

You do realize that this cheat could be used on ANY benchmark out there, including the real-life game-benchmarks? Those benchmarks run a certain fixed demo and record the FPS, just like 3DMark does. If NV cheats of 3DMark, who is to say that they don't cheat on other benchmarks as well? What's different with 3DMark is that it contains tools that can expose this type of cheating (movable camera).

And fact is that 3DMark IS one of the most important benchmarks being used. And NV uses dishonest methods to boost their score. And ome of you are defending them!!!!

Seriously, the bias and blindness of some people just blows me away....

 

[WicKeD]

If a video driver gives me an increase in fps in any game, while not sacrificing the quaility of the images then its all good.

I never benchmark my system by using 3dMArk. Maybe the review sites should stop using this benchmark and just use the current games itself. This would stop the optimizing of drivers for this benchmark. Let the video companies then optimized the drivers for the games that are being benchmarked, that would be a gain for the gaming community.

So in the end, who cares. This is journalisim we are taking about

 

[Yourself]

Originally posted by: WicKeD
If a video driver gives me an increase in fps in any game, while not sacrificing the quaility of the images then its all good.

Pretty much nothing else needs to be said after that. That's the bottom line. If some of you are buying products from a certain company based on the assumption that they are "angels" and wouldn't cheat, you are seriously naive. Buy what runs your games well.


Self

 

[Mem]

TO ALL OF YOU WHO SAY "IT'S JUST 3DMARK, WHO CARES?":

You do realize that this cheat could be used on ANY benchmark out there, including the real-life game-benchmarks? Those benchmarks run a certain fixed demo and record the FPS, just like 3DMark does.

It does say

nVidia believes that the GeForceFX 5900 Ultra is trying to do intelligent culling and clipping to reduce its rendering workload, but that the code may be performing some incorrect operations. Because nVidia is not currently a member of FutureMark's beta program, it does not have access to the developer version of 3DMark2003 that we used to uncover these issues.

Besides when I play games I don`t worry about FPS speed,too busy gaming and it appears to be a FX and 3dmark problem only.

I think it`s getting blown out of proportion IMHO.

 

[Live]

All of you who state you don?t care if they cheat in 3DMark for what ever reason makes me wonder. I agree that 3DMark is not important in itself (at least not to me). What is important to me is my dough. I like to play games and I want them to look and play as good as possible. But on that I have a constraint its technology and most important the money to buy that tech.

So how are benchmarking relevant to tech and your dough? Well is pretty plain to most but I will explain it anyway for you who hasn?t bothered waking up for reality in a while.

The price of a graphics card is not "set" by Nvidia or Ati. It is set on the market. The market is governed by supply and demand. We as the consumers are the demand part. Well basically wherever supply and demand meets is the current price. Ring a bell? Free market, capitalism and such?

The demand for 3D-cards is dependent on a lot of factors. People want different things and play different games and so on. But once you know what you want it?s up to which card you buy that will give you what you want. Here the benchmarking comes in. You could either buy all the cards and test them yourself and see which you like best. Not a very realistic option. So we read reviews which compare the cards to each other. Even if you don?t use reviews or benchmarking to decide on your purchase many others do and the demand for a product is set by all. So others demand decides your price.

You look at different cards from all the manufacturers that you think will meet your standard. Often the standard is performance at a certain IQ level. If one card A produces 100 of performance for 100$ and card B just 90 performance for 100$ at the same IQ level which do you choose? Well if you are a rational consumer you would go for card A. To get you to buy card B they would have to lower the price or raise the tech level and so the performance. This is how it?s supposed to work.

What then if the company selling card B instead decides to just alter the tests so it looks like there card gives more performance for the same dollar?

Well you would get no price cut or better technology to raise the performance level. Even if you see trough all this with your super intellect or what ever, the majority of consumers will not and they basically set the price and raise or lower the performance bar.

So if you want better tech for fewer dollars you should care if Nvidia or any other company cheats. I know I do.

 

[Yourself]

Originally posted by: Live
So if you want better tech for fewer dollars you should care if Nvidia or any other company cheats. I know I do.

I personally don't think it boils down to cheating=more$ for less hardware. People who take the time to discuss this type of thing are not the people who have the impact on hardware prices. We represent such a small piece of the pie and I doubt someone at Dell/Compaq is: 1)Worried about a potential driver cheat or 2)Cares about a 3DMK3 score. If said driver caused system-wide display issues, then you'd have an audience.


Sellf

 

[GTaudiophile]

Beyond3D seems to agree.

 

[WicKeD]

opposite of Beyond 3d Here

 

[GTaudiophile]

Originally posted by: WicKeD
opposite of Beyond 3d Here

2 against 1

 

[WicKeD]

2 vs. 2

add me

 

[Live]

Originally posted by: Yourself

Originally posted by: Live
So if you want better tech for fewer dollars you should care if Nvidia or any other company cheats. I know I do.

I personally don't think it boils down to cheating=more$ for less hardware. People who take the time to discuss this type of thing are not the people who have the impact on hardware prices. We represent such a small piece of the pie and I doubt someone at Dell/Compaq is: 1)Worried about a potential driver cheat or 2)Cares about a 3DMK3 score. If said driver caused system-wide display issues, then you'd have an audience.


Sellf

You obviously did not read all of my post. My point is just as you say that we who take the time to discuss this and can judge a graphic card by other means then bold statements of 30-50% more performance or a single benchmarking program are not the ones that drive demand, an as such the price and as such the amount invested in better technology.

The majority of consumers buying a new graphic card only take a quick look at the headlines on the sites and in the magazines before they buy. And if it states this card gets a lot o improvement in the tests we have made and we recommend you buy it, they will.

And they set the price for you!

 

[Yourself]

Originally posted by: Live


The majority of consumers buying a new graphic card only take a quick look at the headlines on the sites and in the magazines before they buy. And if it states this card gets a lot o improvement in the tests we have made and we recommend you buy it, they will.

And they set the price for you!

.....actually, ....no they don't set the price.


Self