At this point, who can trust nVidia?

AMD_Man · May 24, 2003

This is a very interesting story...nVidia is cheating and they're denying it although FutureMark clearly says they are...it makes you wonder about other benchmarks that suddenly increase in performance due to the Detonator FX drivers. Who is to say that nVidia didn't cheat in other games as well?

Intelligence is not merely the wealth of knowledge but the sum of perception, wisdom, and knowledge.

dhlucke · May 24, 2003

Trust no one. You're not safe.

Is Nvidia cheating?

<A HREF="http://www.extremetech.com/article2/0,3973,1086025,00.asp" target="_new">Extremetech says YES!</A>
<A HREF="http://www.theinquirer.org/?article=9648" target="_new">Futuremark says YES!</A>

TKH · May 24, 2003

Yes, trust nobody even yourself because sometime you don't know what you are doing... and don't trust me...

It is morally wrong to allow suckers to keep their money
<A HREF="http://www.anandtech.com/mysystemrig.html?id=22996" target="_new">My System Rig</A>

noko · May 24, 2003

I wouldn't take it from one extreme to another. A level of trust should be establish before you make a buying decision. You don't want to buy a $400 video card to open up a box with crap in it, would you? You would at least expect to get a video card, a video card that is hopefully represented on the box, a video card that works etc.. Right? Nvidia is clearly misleading a number of folks indirectly you can say to believe something about their product which it is not. Is that a trust issue, you bet. Would I buy a Nvidia product right now? Maybe but it wouldn't be a video card I dare say.

jaythaman · May 24, 2003

Whats really the pits is that a lot of people and i mean a LOT will end up buying the Fxs just because of this. I dont really know many people who do much internet research or post in forums for advice. They will fall for this fake PR. Look at the amount of fx5200s that have been selling and the amount of people asking advice for a choice between 5200s and ti 4200s. I mean is there even a comparison?

What really bugs me is that why does nVidia has to stoop to such low levels ....They have produced wayy better products in the past and has been the hands down market leader for a long while. Look at nForce IT's damn good and the only choice for AMD procs. Can't they just get back to some serious work? IT's desperation at it worst
OKK!! Which one of you wise guys stole my sig! :frown: :tongue: Jay Kay

Guest · May 24, 2003

The FX5900 outperforms the R9800 in Tom's HW game benchmarks and that's all I care about.

--Why do blondes have bruises on their belly buttons? Because blonde GUYS can be dumb too! Ha ha ha ha ha ha ha!!!!!--

AMD_Man · May 24, 2003

The FX5900 outperforms the R9800 in Tom's HW game benchmarks and that's all I care about.

So what? That means nothing if nVidia is cheating to get those scores.

Intelligence is not merely the wealth of knowledge but the sum of perception, wisdom, and knowledge.

Sarke · May 24, 2003

So what? That means nothing if nVidia is cheating to get those scores.

Let's not get carried away. I know nVidia cheated (yes, it's cheating, but so did ATI and ATI is just glad that they didn't do it as extremely as nVidia), but that was only in one specific benchmark. They didn't cheat in the games, and they could still beat the 9800 (not by much, but still). So just because they cheated in one single benchmark doesn't mean that their entire product is made of air (no pun intended). It is still a good card and and is still comparable to what ATI offers.

coolsquirtle · May 24, 2003

poof cheating poof~~~~~ big deal~~~ only thing i care about is getting more then 2fps when i get doom3. and no whoever can tell the difference between 45fps and 48fps in any game should be shot....... why are you people making this such a big deal anyways..... so they cheated BIG DEAL! anyone who said they have never cheated or help someone cheat should be shot in the head as well, also i think personally that the only reason u guys are making a big fuss about this is you ATi Fanboys, scared that FX5900Ultra beat 9800pro.............what was it? 800 mark that nvidia cheated off?? tell me does that really matter -_-" as long as i can get more fps in Unreal 2k3 i don't really care how they cheated on a benchmark program infact i doubt among the people that will be a FX5900Ultra or 9800pro probably half of them don't even know wtf benchmark is

Proud Owner the Block Heater
120% Nvidia Fanboy
Athlon 64......nForce 3, it sure sucks when you can't afford things. One day Operton Servers will cost less than $4000 so everyone can get one

Twitch · May 24, 2003

And if you weren't "120% Nvidia fanboy," you would feel the same way? I always find it hilarious when one fanboy starts calling other people fanboys.

The Netherlands is where you go when you're too good for heaven. :tongue:

simwiz2 · May 24, 2003

The game benchmarks are all that really matter. I don't care much about these benchmark programs, and having both companies cheating in a single round of benchmarks just proves that I have good reason to completely ignore 3DMark.

eden · May 24, 2003

No, it proves that companies are immature and are forgetting to optimize Drivers for games, and instead for benchmarks, when these are used to TEST a card, to check if its FEATURES WORK.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:

Guest · May 24, 2003

What he said.

Nvidia didn't cheat on the games.

I'm willing to live with a superior GPU, even if it's from an untrustworthy company. Games are all I care about, 3D Mark has always been controversal anyway.

--Why do blondes have bruises on their belly buttons? Because blonde GUYS can be dumb too! Ha ha ha ha ha ha ha!!!!!--

dhlucke · May 24, 2003

What it proves is that Nvidia's product wasn't performing as well in 3dmark03 as they hyped it so they cheated.

If Nvidia didn't overhype and paper release their GFX cards, they wouldn't need to cheat. They refuse to settle for second best.

Is Nvidia cheating?

<A HREF="http://www.extremetech.com/article2/0,3973,1086025,00.asp" target="_new">Extremetech says YES!</A>
<A HREF="http://www.theinquirer.org/?article=9648" target="_new">Futuremark says YES!</A>

Twitch · May 25, 2003

The GPU may or may not be superior, but is it worth $100 more? That's why nVidia is cheating. It's hard to ask an additional $100 for a card that has 3-5% better performance overall. Much easier if you can claim 15-20% performance boosts...

And incidentally, calling the NV35 GPU "superior" is very debatable. It is a tiny bit faster, nothing more. I suppose if FPS is your only criteria to define quality in a graphics card, it's somewhat superior. There's more to 3D graphics than framerates, however.

The Netherlands is where you go when you're too good for heaven. :tongue:

Edited by Twitch on 05/24/03 05:21 PM.

Guest · May 25, 2003

Can't argue that. The FX5900 is a big rip-off compared to the R9800, and especially the R9700. I'm not buying a GPU until July/August, so I'll just have to see how price are by then. If the FX5900 is $50 more expensive, sure I'll buy it, but right now, you're 100% right.

--Why do blondes have bruises on their belly buttons? Because blonde GUYS can be dumb too! Ha ha ha ha ha ha ha!!!!!--

eden · May 25, 2003

Not just that, but the 5800 is slightly behind it in most standard tests and even sometimes ahead, such as in Splinter Cell!
Is it worth the cost?
Imagine how worth it really is, for a 256MB version at 499$, insane pricing.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:

Spitfire_x86 · May 25, 2003

only thing i care about is getting more then 2fps when i get doom3

OK, now go and play all games @ 320 x 240 or the lowest res. possible and least details to get the best fps

----------------
<A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A>

<A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A>

Crashman · May 25, 2003

Great! So I should buy an nVidia card because they cheated on a benchmark, so I can feel sneeky. And a P4X400 board because VIA stole that, even though they later were forced to pay. And I'll be running some copy of Windows XP I picked up off Kaza. Now if I could just find a way to run that XP3200+ on it, and overclock it to the point that it actually performs on par with a P4 3.2C.

Watts mean squat if you don't have quality!

rain_king_uk · May 25, 2003

How do you know they didn't cheat in the game benchmark's too?

Sarke · May 25, 2003

How do you know they didn't cheat in the game benchmark's too?

Cuz the only way they could cheat in a game is if they reduce the image quality and ppl would notice that. But if they do write game-specific code then that is an optimization: game runs faster and looks the same, therefore not a cheat.

eden · May 25, 2003

The new problem arising in optimizations, is that it nevers used to be the driver that created performance increases, it used to be game-specific, as in THE GAME ITSELF is programmed. When Carmack does Doom III, HE is the one who optimizes, not ATi. HE makes sure the game will run at its best on ATi. Nowadays it seems most other programmers are lazy, and instead rely on driver control, which btw is NOT supposed to be this relied on, because the main goal of drivers, has and always will be, functionality.
Any performance boost aside, is a boon. Giving us this boon due to laziness, IMO, just isn't fun. And can you imagine how hard and long it is, to make game-specific opts in drivers? Pretty soon, we will download drivers 100MBs big, because the company wanted a leg up over the other in games.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:

Twitch · May 25, 2003

How many of the specific game benchmarks that we see are really demos being run on rails? Because it seems like any such benchmark might be susceptible to predictive culling. And that's not an optimization, it's a cheat. It's only an optimization if it carries over into gameplay, and a graphics card can never predict which way a player is going to turn the mouse...

The Netherlands is where you go when you're too good for heaven. :tongue:

rain_king_uk · May 25, 2003

Actually I think since the game benchmarks are pretty much preset also, there are plenty of ways to cheat in them that are nothing to do with optimizing the game.

I would place a bet that both nVidia and ATi will be worrying right now about what they are going to do about cheats they have used in their drivers for game demos that are widely used for benchmarking. They can't just take them out in future drivers because it will be obvious.

I think it's only a matter of time before more cheating comes to light.

GeneticWeapon · May 25, 2003

Well, it seems as though Nvidia has been caught cheating in the SplinterCell benchmark also....that is being uncovered right now as we speak....I dunno, they're really poo poo heads.

3DMark 03 = 4,101
<A HREF="http://service.futuremark.com/compare?2k3=775464" target="_new">http://service.futuremark.com/compare?2k3=775464</A>
"I want that dam Cramack to finish Dom3 Now I wanna play!" - young rage3d forum member

At this point, who can trust nVidia?

Splendid

Polypheme

Distinguished

Distinguished

Distinguished

Guest

Guest

Splendid

Distinguished

Distinguished

Distinguished

Distinguished

Champion

Guest

Guest

Polypheme

Distinguished

Guest

Guest

Champion

Splendid

Polypheme

Distinguished

Distinguished

Champion

Distinguished

Distinguished

Splendid

Share this page