More Nvidia cheating ? 3Dmark2001 now!

[RobsTV]

Blastman
Thanks for the firing squad link.
It's nice to see benchmarks for games that "I" actually run.
And after seeing the results in Nascar 2003 I'm glad I stuck with nVidia for my machine, and left the ATi card for the wife to game with.
Yes, even firingsquad shows the 4200 spanks the 9800 badly in this game 4200 60% faster than 9800 Pro at 1024x768x32

But that is drifting off the thread topic.
Topic is 3dm2k1se, and as proven, if anyone cheated at that it was ATi.

EDIT: it seems that since my initial post above, firing squad has changed the results from what they posted earlier. Here is what they say the reason was:

"After countless hours troubleshooting the 875P/RADEON combination with CATALYST 3.4 with no success, we finally decided to try an nForce2 setup. Voila, the RADEON scores jumped significantly! "

So in order for Firing Squad to post an ATi score high enough to beat NVidia, they had to use a NVidia based motherboard. How ironic is that?

In any event, the original page they had up that showed NVidia faster has now been replaced with this "modified for this test" page, so my remarks in inital post above are no longer valid.

If you look at their test setup, they now show:

Intel Pentium 4 3.0GHz (Hyper-Threading enabled)
AMD Athlon XP 3200+ (for Nascar system)

ASUS P4C800 Deluxe (875P, Turbo mode enabled)
ASUS A7N8X Deluxe II (for Nascar system)

So all credibility in their testing methods has gone out the window.
I could just imagine what they were thinking.........
Hey, we don't like the results on one test, so let's change motherboards until we get the results we want!
Then we won't bother rerunning all the other 5 tests with this motherboard, since they had the results we were looking for.

 

[Blastman]

And after seeing the results in Nascar 2003 I'm glad I stuck with nVidia for my machine, and left the ATi card for the wife to game with.

Sorry, better think again. They ran the scripted demo. Again, it?s a scripted benchmark like the ones Nvidia is cheating on.

The 4200 was 31% slower in Nascar Racing 2003 than a 9500pro ? here?in HardwareFrance?s review. Needless to say a 9700pro would be a lot faster yet.

Do you now see the problems we are having because of Nvidia?s cheating? Which card do you choose now? Who?s benchmarks do you trust?

 

[Rand]

Originally posted by: ketchup79
Can you think of any game where the 8500 is even close to the speed of a ti4200? No? Then these adjusted scores are invalid in my book.

Yes, I can think of 5 of the top of my head. I can also think of 3 applications (Non-GLide) in which the V3 outperforms the R9700 Pro.
With a broad enough data set it's not hard to find driver issues/weaknesses in virtually any graphics architecture relative to another.
One can find multiple areas in most any individual game in which the R88500/GF4 Ti4200 may switch the lead.

You'd be hard pressed to find any graphics chip that is so perfectly well rounded and clearly superior that it beats another in every respect under any possible conditions regardless of optimizations.

I have little doubt that given enough effort one could find individual scenarios in games on modern systems in which the GF4 would be hammered badly by the classic TNT.

 

[classy]

Originally posted by: nippyjun

Originally posted by: peter7921
Ever since nVidia acquired 3DFX it has turned for the worst. Products are becoming more and more 3dFX like(big huge cards that are over-hyped and under-performing) and ethics are becoming more and more 3DFX like. It looks like 3DFX conqured nVidia from within!!!

Have a little respect for 3dfx. After all they revolutionized gaming on the pc.

Agreed and 3dfx never was caught cheating. And ATI's drivers have been pretty much "optimized free".

 

[classy]

Originally posted by: Rand

Originally posted by: ketchup79
Can you think of any game where the 8500 is even close to the speed of a ti4200? No? Then these adjusted scores are invalid in my book.

Yes, I can think of 5 of the top of my head. I can also think of 3 applications (Non-GLide) in which the V3 outperforms the R9700 Pro.
With a broad enough data set it's not hard to find driver issues/weaknesses in virtually any graphics architecture relative to another.
One can find multiple areas in most any individual game in which the R88500/GF4 Ti4200 may switch the lead.

You'd be hard pressed to find any graphics chip that is so perfectly well rounded and clearly superior that it beats another in every respect under any possible conditions regardless of optimizations.

I have little doubt that given enough effort one could find individual scenarios in games on modern systems in which the GF4 would be hammered badly by the classic TNT.

This is also true. I do doubt that TNT classic scenario though. But I do agree with you.

 

[MercenaryForHire]

Originally posted by: classy

Originally posted by: nippyjun

Originally posted by: peter7921
Ever since nVidia acquired 3DFX it has turned for the worst. Products are becoming more and more 3dFX like(big huge cards that are over-hyped and under-performing) and ethics are becoming more and more 3DFX like. It looks like 3DFX conqured nVidia from within!!!

Have a little respect for 3dfx. After all they revolutionized gaming on the pc.

Agreed and 3dfx never was caught cheating. And ATI's drivers have been pretty much "optimized free".

<Duck> Quack. </Duck>

- M4H

 

[Blastman]

RobsTV ?. , here is another Nascar benchmark (2002) by Extremetech? ? one of the independent benches they developed for their 3DGameGage. The 5900U gets beat handily (45%) by the 9800pro.

Those FiringSquad Nascar benches are literally ridiculous. They either messed something up in the settings bigtime or Nvidia?s drivers are cheating bigtime. I simply don?t believe those scores. The Extremetech benches correlate well with the HardwareFr review and suggest the ATI cards are much faster in Nascar.

 

[RobsTV]

Originally posted by: Blastman
RobsTV ?. , here is another Nascar benchmark (2002) by Extremetech? ? one of the independent benches they developed for their 3DGameGage. The 5900U gets beat handily (45%) by the 9800pro.

Those FiringSquad Nascar benches are literally ridiculous. They either messed something up in the settings bigtime or Nvidia?s drivers are cheating bigtime. I simply don?t believe those scores. The Extremetech benches correlate well with the HardwareFr review and suggest the ATI cards are much faster in Nascar.

Look, you can't pick and choose results like you are doing.
Can't take the results you want from a site, then throw out the others at the same site.

First you say, 3dm2k1se shows NVidia is cheating. You then go on to say 4200 actually runs like a 8500.
Then you are shown Tom's VGA charts that clearly show only ATi could be cheating in 3dm2k1se.

Next you say, but look at Firing Squad for better reviews than Tom's, because they don't use scripted demos.
Where did your 4200 Vs 8500 comparison go????
Heck Firing Squad thinks so little of the 8500 they left it out of benchmarks. ( I myself think the 8500 is a great card)
Then you are shown results at Firing Squad of a non scripted test, Nascar Racing 2003.
Demo means not the full game, but simply a demo which only includes one track, Daytona.
Nothing graphical is "fixed in place" so there is no way to know what a sceen will have present in it ahead of time.
You have full control over every setting, just like the full version, only limited to one track.
The actual test was made up by Firing Squad, and now you say Firing Squad sucks because those results are not what you want..

Next you link to French site that shows the 4200 is faster than a 5600? And you think they are accurate?????

Then you link to another site and say but look at the 5900 Vs 9800, in tests with Nascar 2002.
What happened to testing Nascar 2003? What, they couldn't afford new games to test on?
I think there are many like myself who could care less about which top dollar card is best, 5900 or 9800
and how it relates to 3dm2k1se. You don't get those expensive cards to run this thread title.
Where did your 4200 Vs 8500 comparison go???? You remember them? The cards made to run 3dm2k1se?
Why drift off track in your comparison of cards to now bring up 5900 and 9800, in old games?
Show us proof the the 8500 and 4200 run the same, and that NVidia is cheating in 3dm2k1se.

Looks like your agenda is NOT 3dm2k1se cheating, but simply more NVidia bashing.
Give it up, you are beating a dead horse with the 3dm2k1se story.
But good try to recover though.

 

[chsh1ca]

Originally posted by: RobsTVYes, even firingsquad shows the 4200 spanks the 9800 badly in this game 4200 60% faster than 9800 Pro at 1024x768x32

LOL, try reading your link again.

Topic is 3dm2k1se, and as proven, if anyone cheated at that it was ATi.

Proven? Did I miss something? Where is the proof?

Look, you can't pick and choose results like you are doing.
Can't take the results you want from a site, then throw out the others at the same site.

Yes you can, especially if the results were obtained using different methods. The goal here was to utilize non-scripted benchmarks (AKA, user-made demos) in order to test various games, just to see how well the cards really do perform, rather than how well they perform on those particular demos. We know nVidia has been caught once heavily optimizing their drivers for specific benchmarks that use 'scripted' demos and we also know that ATI was caught pulling some tricks behind the scenes on the very same demos. It is in everyone's best interest to get the REAL numbers.
In fact, it would be bad science to just simply use the demos everyone else is, especially when your goal is to present and provide supporting proofs for a hypothesis that disagrees with general consensus (which in and of itself is rapidly changing now).

First you say, 3dm2k1se shows NVidia is cheating. You then go on to say 4200 actually runs like a 8500.

Is that an overly inaccurate statement?

Then you are shown Tom's VGA charts that clearly show only ATi could be cheating in 3dm2k1se.

Given that there is evidence to the contrary, I don't believe that to be the case. It may help if futuremark were to release a patch for 3dm2k1se as well, then we'd have a clear indicator of who's doing what funky business. Another problem that also rears its head is that the THG article quoted uses rather old drivers (28.32 and 6.1x), and we know that there have been several improvements made along the way.

[...]

Next you link to French site that shows the 4200 is faster than a 5600? And you think they are accurate?

Where's the proof that it is inaccurate? You've been presented with evidence that states in certain games, the Ti4200 is faster than an FX5600U (Something Splinter Cell already showed us), and you disagree with the results? You seems to have delusions about how powerful the FX5600U actually is.

Then you link to another site and say but look at the 5900 Vs 9800, in tests with Nascar 2002.
What happened to testing Nascar 2003? What, they couldn't afford new games to test on?

CounterStrike, last I checked, is still the most played online game. Perhaps we should just ignore the bugs with it in the recent driver releases then, since it's an old game? No, we shouldn't. It's good to know how the cards will perform on all games, regardless of age. Especially if they are extremely popular.

Cards are not bought to play games that are coming. If someone tells you that, they are lying. If ATI or NVIDIA released a card that pulled 5FPS in UT2003@1600x1200x32 4xFSAA/8xAF, but was capable of 120FPS in Doom 3 at the same res, would you buy it? Of course not. People want immediate return on what they pay for.

 

[squidman]

Lol, i was on blastmans side the whole time...but yo boy is just overreacting. Freaks me out - so much energy wasted just to type what chaos has sinister nVidia (or Ati) brought upon us! oh horror!
Reminds me of whole spanish-soap-opera craze here in Moscow: "Did u see what Roderigo did to Luisiana?" "yes, hes a bastard to bertray her like that" "I think Jose is older than he ios pictured in the movie" "Oh, yes, i think Jose is well over 40! Yet in the movie, he says hes 25!" "cheats!" "Yes, o horrors!".

lol

 

[Blastman]

Well, I was just about to post my analysis of why FiringSquad?s Nascar 2003 graphs were wrong, and they just updated them within the last hour. I figured they were wrong. They originally had a 9800pro 60% back of a 4200.

Here is my analysis anyway because we need to scrutinize benchmarks and analyze them.

Fist point.

Do the graphs make sense? NO. Look at the Nascar 2003 graphs and what do you see? You see the 4 Nvidia cards (4200, 5800U, 5900U 5900) all producing the same score at 10x7. This means the game is CPU limited at that that resolution -- a 4200 can match a 5900U because of it. The high end ATI cards (9700pro, 9800) ? ?should also? ? be scoring at that same level of the Nvidia cards if the game is easily CPU limited at that resolution. But they are not --- if fact they?re 60% back which is ridiculous --- ergo --- conclusion ? FiringSquad really messed up the settings for the ATI cards. So just by looking and analyzing the graphs one can conclude FiringSquad messed up.

Second point.

The FiringSquad benches do not correlate at all with the 2 other Nascar benches I produced. In fact they?re way off. Again this points to the conclusion that FiringSquad messed up their numbers. It may have been a cheat but I tend to the conclusion that FiringSquad messed up the settings for the ATI cards.

Note: Even though one of those 2 benches was Nascar 2002 and not 2003, I think it?s a reasonable assumption that the game engine is relatively the same for the 2 versions and that they should produce very similar results.

As far as the Nascar AA/AF benchmarks. I?m quite sure there is something wrong there too. Hopefully FSQ will check and update those too.

 

[squidman]

goddamn, its just o mofin video card. I dont even spend this much energy on my girlfriends, let alone some "nVidia cheated" thread. Gotta kick myself in the nuts for this. *smacks* ouch!

 

[squidman]

I see the tensions are growing, so heres a parody post. Its evening, im kinda drunk from a good day at friends house, so here goes.
*The nVidia's cheat, or how Ati remained the only honest videocard*

(played as a spanish soap opera)

Blastman = Roderigo
RobsTV = Old Giovanni
nVidia = sexy girl ¹1
Ati = sexy girl ¹2.

Roderigo: NO GIOVANNI! I WONT ADMIT IT!!! nVidia cheated me!!! She cheated us all while she was at it!
OG: Come on, Roderigo...she didnt cheat you. She did nothing wrong! Shes a sexy young woman, and she craves attention. And she gets it like smart old men like me *pats his groin*.
R: OLD GIOVANNI! HOW COULD U DO THAT?!
OG: Do what? ohhh...do you know how many people did her?! haha! i was just the one who enjoyed our *experience* so much, i didnt go around talking $#!+ about her!
R: so....maybe what i did is wrong?
----end of act 1-----
----act 2----
OG: *twirls his mustache* that nVidia girl cando wonders...how come she got such measly reviews?
R: AHA!!!! See how many people HATE her? You thought i was crazy, haha! It is not just i, it is the Fire Squad, the Extrem Tech, and other brothels where she was attending!
OG: nonsense, boy. Why do you only focus on bad reviews? I got a good experience, and shall always put in a good word for nVidia.
R: well, u may be right...but nVidia is NOTHING when compared to Ati! Yesterday, i borught her flowers!
OG: Flowers?! *chuckles*. I did her yesterday, while my wife wasnt home.
R; *gasps* U *DID*??!
OG: yeah, so?! I mean shes cool, satisfies me, but i still prefer nVidia - shes the hotest.
R: *sobs* bertrayal....
OG: And i really dont understand, if theyre are about the same - how come Ati is always preferred to nVidia....hmmm....
R: *to himself*: maybe because nVidia likes big peni*ses.
OG: SO THATS WHATS UP!!! Haha! Now its become clear! Fools just have to grow dicks and wisdom to undersatnd and enjoy nVidia just like i did! *points finger* haha! u got a small one!
R: I shall take this no longer!!! I give up my sexuality, and convert to homosexualism! That new pool-cleaner Matrox seems like an excellent man! I shall love HIM! *cries out, and storms outta the room*
OG: *chuckles* and joins nVidia and Ati in the hot tub.
nVidia and Ati in unison: we do like a strong male!!!!
OG: and a *big* one at that!
Everybody: hearily laughs.


THE END

 

[Schadenfroh]

lol

 

[chsh1ca]

Originally posted by: chsh1ca AT 10:13 AM

Originally posted by: RobsTV
Yes, even firingsquad shows the 4200 spanks the 9800 badly in this game 4200 60% faster than 9800 Pro at 1024x768x32

LOL, try reading your link again.

Originally posted by: Blastman AT 2:31 PM
Well, I was just about to post my analysis of why FiringSquad?s Nascar 2003 graphs were wrong, and they just updated them within the last hour.

 

[MercenaryForHire]

Squidman rating ++

- M4H

 

[nemesismk2]

Originally posted by: classy

Originally posted by: nippyjun

Originally posted by: peter7921
Ever since nVidia acquired 3DFX it has turned for the worst. Products are becoming more and more 3dFX like(big huge cards that are over-hyped and under-performing) and ethics are becoming more and more 3DFX like. It looks like 3DFX conqured nVidia from within!!!

Have a little respect for 3dfx. After all they revolutionized gaming on the pc.

Agreed and 3dfx never was caught cheating. And ATI's drivers have been pretty much "optimized free".

3dfx never had to cheat because pretty much everything was optimised for their video cards anyway.

 

[nemesismk2]

NVIDIA CHEATED (MAYBE), OH NO IT'S THE END OF THE WORLD!

I CAN SEE THE TABLOID HEADLINES NOW "SAD ANANDTECH GEEKS LOSE THEIR MINDS OVER ALLEGED NVIDIA CHEATING!"

 

[chsh1ca]

Originally posted by: nemesismk2
NVIDIA CHEATED (MAYBE), OH NO IT'S THE END OF THE WORLD!

I CAN SEE THE TABLOID HEADLINES NOW "SAD ANANDTECH GEEKS LOSE THEIR MINDS OVER ALLEGED NVIDIA CHEATING!"

Yep, just the same response that would exist if it were discovered that ATI was cheating as grossly as NVIDIA has been.

 

[Ketchup]

Originally posted by: chsh1ca

Originally posted by: nemesismk2
NVIDIA CHEATED (MAYBE), OH NO IT'S THE END OF THE WORLD!

I CAN SEE THE TABLOID HEADLINES NOW "SAD ANANDTECH GEEKS LOSE THEIR MINDS OVER ALLEGED NVIDIA CHEATING!"

Yep, just the same response that would exist if it were discovered that ATI was cheating as grossly as NVIDIA has been.

Anybody remember the incident with Quake 3 and the Radeon 8500? If you don't, you can bet everyone will forget about this incident in about a year. Get over it already.