nVidia Responds "It was a right-wing conspiracy" + ATI's Response

[WetWilly]

From nVnews:

NVIDIA has issued a brief response in regards to the alleged cheating that FutureMark claimed took place in certain Detonator FX drivers.
It is not known at this time if NVIDIA plans to issue further statements on the matter.

"Since NVIDIA is not part in 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. 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. This is obvious since our relative performance on games like Unreal Tournament 2003 and Doom3 shows that The GeForce FX 5900 is by far the fastest graphics on the market today."

ROTFLMAO!!! What's the matter? Did Jen-Hsun Huang make Bill Clinton nVidia CEO-for-a-day?

 

[WetWilly]

Here's ATI's official response:

"The 1.9% performance gain comes from optimization of the two DX9 shaders (water and sky) in Game Test 4 . We render the scene exactly as intended by Futuremark, in full-precision floating point. Our shaders are mathematically and functionally identical to Futuremark's and there are no visual artifacts; we simply shuffle instructions to take advantage of our architecture. These are exactly the sort of optimizations that work in games to improve frame rates without reducing image quality and as such, are a realistic approach to a benchmark intended to measure in-game performance. However, we recognize that these can be used by some people to call into question the legitimacy of benchmark results, and so we are removing them from our driver as soon as is physically possible. We expect them to be gone by the next release of CATALYST."

 

[oldfart]

This is obvious since our relative performance on games like Unreal Tournament 2003 and Doom3 shows that The GeForce FX 5900 is by far the fastest graphics on the market today."

Looks pretty darn slow in UT 2K3 in these [ H ] screenshots. 1/2 to 1/3 the FPS of a 9800P. Whats the deal?

FX 5900U 8xAA/8xAF UT2K3 ---37 FPS
9800P 6xAA/16xAF UT2K3 ---95 FPS

FX5900U 8xAA/8xAF UT2K3 ---30 FPS
9800P 6xAA/16xAF UT2K3 ---69 FPS

 

[WetWilly]

I know - after reading all these FX5900 Ultra reviews, I'm still trying to figure out what's the BFD for $499? The card's not out yet, benchmarks are far from decisive, Doom3's not out yet and ATI's got plenty of time for a driver refresh for Doom3 and the 9x00 cards, and the 5900 is fawned over.

Looking at the big picture, I think this is proving ATI got a much better deal buying ArtX than nVidia did buying 3dfx.

 

[Genx87]

Take a look at the comparison ROTFLMAO

8XAA vs 6xAA and you think that is valid? Apples to oranges.

Gen

 

[oldfart]

I see that you forgot to mention nV 8x AF Vs ATi 16x AF. This was "max IQ" settings for both cards. Answer this. Which looks better? Thats right. The 9800P 6x AA looks BETTER than nV 8X AA (unless you LIKE jaggies). You think 6x AA to 8x would make that much of a difference? The 9800P in this test looks better and has WAY better FPS. I'll take 90 FPS over 30 any day thanks.

 

[kylebisme]

This is obvious since our relative performance on games like Unreal Tournament 2003 and Doom3 shows that The GeForce FX 5900 is by far the fastest graphics on the market today.

lets see:

1) the radeon comes out on top ut2003 from what i have seen, as mentioned above.
2) doom 3 isn't even out so it does not rightly fall into the category of "today".
3) the fx5900 is not on the market, so that does not rightly fit into the category of "today" either.


is nvida smoking something hallucinogenic?

 

[nRollo]

LOL You weasels are just trying to make yourselves feel better that you're not going to part with $500 to have an FX 5900 Ultra, the fastest gaming solution on the planet by far.
Old Fart, the 9800 Pro wins some ultra high res/AA/AF benches? Big deal. They are settings NO ONE WOULD RUN BECAUSE THE ONLY THING WORSE THAN JAGGIES IS STUTTERING!

 

[oldfart]

No one with a 5900U would run it @ 30 FPS. R9800 owners can run it just fine @ 90 FPS and have better IQ.

Ultra high res with great IQ, max AA, and AF and smooth framerates is what I expect from a top of the line card. Not 30 FPS stuttering with jaggies.

 

[Rogozhin]

The 5900 is not worth the $500 no matter how you look at the benchmarks.

rogo

 

[Shamrock]

isnt AA more stressful on the GPU than AF is? because the NV has 8x and the ATI has 6x? just a thought.

I know I can run 4xAF on my GF3 Ti200 and the FPS will only drop 6-7, but if I enable 2xAA on it, I get reduced by 20-30 fps!

 

[oldfart]

No matter how you look at it, the 9800P max AA is better looking (see the jaggies on the 8X AA 5900U?) than nVidia's. The framerate difference when at max quality is huge One is very playable, the other, not at all. If I plunked down a wad of cash for a card like that, cranked up the detail and got 30 FPS, I would not be very happy. My Ti4600 was like that. Played great until I turned up the AA/AF. The performance dropped like a rock.

 

[Rand]

And here I was thinking this might be a discussion of the corporations responses rather then an argument over which graphics card was superior.... silly me.

 

[Pete]

Shamrock, who cares how stressful something is on a GPU? Do you feel sorry for the 8500 when it's compared to the GF3/4 in AA benches, though the former uses SSAA and the latter, MSAA? All customers care about is how fast a card renders at a certain IQ compared to other cards in teh same price class. ATi currently has nVidia beat in terms of AA IQ and performance (at comparable IQ--keep in mind nV can't duplicate ATi's gamma corrected AA).

Rand, what's there to discuss? nVidia is taking the low road again, with more FUD and, IMO, lies. 5900U doesn't win UT2K3 as handily as nVidia claims it does, and neither Doom 3 nor the 5900U are on the market yet. And the only one who made nVidia's FX series look bad at floating point shader rendering is nVidia, not FM or ATi. If nV thinks 3DM03 is such a useless test, why don't they just say so and not optimize for it at all? I don't see how trying to trick everyone and then lying about it is easier or more beneficial than educating every webiste they give a review card to about how 3DM03 isn't very relevent in the first place (and I think they could have done so fairly successfully, showing how well they do in most game benches, and how nV is still the developer standard).

They're spreading FUD just like they did with the Kyro II and cheating like with 3DM01, and it's disgraceful.

 

[Painman]

nVidia: "Futuremark is out to smear us. Screw the 3DMark numbers. Have a gander at these UT2K3 and DOOM 3 numbers!"

ATi: "We applied an optimization technique to the benchmark which does not impact IQ and is legitimate in a game-specific case, but may not be in a benchmark-specific case. We're getting rid of it."

Do I expect corporations big and small not to cheat?

Hah, I wasn't born yesterday, folks.

Do I expect corporations to employ diversionary tactics when caught with their pants down?

See above.

Am I pleased when I see corporations accept some kind of accountability for their actions and offer remedies for them?

I'm annoyed that they went and finagled in the first place, but it's nice to see them own up to it.

Are synthetic and/or pre-fabricated benchmarks a lousy and exploitable way to gauge real world performance?

Yes.

Doesn't that make this entire fiasco moot?

No.

Why not??!?

Because the stupid, through their sheer juggernaut-like force, have assumed control, and the question of whether these benchmarks are truly meaningless, is what is moot. These benches are what they pay attention to, and are that upon which they stake the fate of hundreds of millions of dollars.

ATi and nVidia both know this, and so should you.

<Morpheus> Welcome... to the Real World. </Morpheus>

 

[kylebisme]

Originally posted by: Rollo
LOL You weasels are just trying to make yourselves feel better that you're not going to part with $500 to have an FX 5900 Ultra, the fastest gaming solution on the planet by far.
Old Fart, the 9800 Pro wins some ultra high res/AA/AF benches? Big deal. They are settings NO ONE WOULD RUN BECAUSE THE ONLY THING WORSE THAN JAGGIES IS STUTTERING!

lol, i will part with 500$ to have a card that is considerably faster than my 9700pro when such a thing becomes available and when i actually need something faster. however, if i were to need more speed today i would go for the 9800pro which can pump out higher image quality than a 5900u and do so AT FRAMRATE FAR ABOVE STUTTERING!

it is funny Rollo, you were saying that nvidia had the best card around on their last try which they have sense dumped because it was not nearly what they or you were trying to make it out to be, now history is repeating itself. maybe if you stick to your guns and keep it up long enough you will eventually be right.

 

[Lonyo]

Originally posted by: Rollo
LOL You weasels are just trying to make yourselves feel better that you're not going to part with $500 to have an FX 5900 Ultra, the fastest gaming solution on the planet by far.
Old Fart, the 9800 Pro wins some ultra high res/AA/AF benches? Big deal. They are settings NO ONE WOULD RUN BECAUSE THE ONLY THING WORSE THAN JAGGIES IS STUTTERING!

You sir, are a 'tard.

If I can waste $500 on a graphics card, I would almost certainly have a very impressive system to go with it (possibly 3GHz processor, 1GB RAM and all the trimmings).

With this setup, I would probably also have a nice big monitor (18" LCD/19 or 21" CRT) and I would want to use it to its full potential.

Thus, with my shiny, fast new system, I would want some damned high settings on all my games. If I have a very expensive system, I would want to be able to play games with all the extra bits on and everything up high.

Thus, a top end graphics card that costs a lot MUST be good at the top end settings that make the game look a lot better, or they are, in fact, not all that great.

Maybe this will not be true of all users who buy such $500 cards, but I would guess that most people who have that money to spend would want the highest setting, and want them to run fairly well. They would expect speed and IQ.

The high AA/AF settings are exactly what the people who would buy these cards would run, unless they were silly and were happy playing 1024x768 32bit colour medium details on their new machine.

 

[BenSkywalker]

No matter how you look at it, the 9800P max AA is better looking (see the jaggies on the 8X AA 5900U?) than nVidia's.

As long as you don't look at games that use alpha textures that is. The 8x mode is running SSAA and MSAA offering a vastly superior solution for older games then anything on the R3X0 core boards. Then you also have the issue of texture aliasing, ATi has plenty of it, enough for them and nVidia to share which is good as running 8xs you don't have any noticeable at all on the nV part

I'm considering the 5900Ultra at this point over the other models mainly due to older games and 8xs(the frame buffer requirements make it quite a stretch for a 128MB board running @res with 8xs).

The 5900 is not worth the $500 no matter how you look at the benchmarks.

And the R9800Pro 256MB is?

 

[nRollo]

Lonyo:

You sir, are a 'tard.

What a pleasant thing to say Lonyo. You are well mannered indeed.

If I can waste $500 on a graphics card, I would almost certainly have a very impressive system to go with it

It would be foolish to put a card like that on your computer Lonyo, it would be worth more than the computer. I do have a P4 2.53/Asus P4PE/512 PC2700/21" CRT though, so I think I'll buy one. It will be pretty close to what you envision as the "appropriate" system, and I'll upgrade my mother board by year end in any case.

"Thus, with my shiny, fast new system, I would want some damned high settings on all my games. If I have a very expensive system, I would want to be able to play games with all the extra bits on and everything up high."
You go Lonyo. BTW- why is it you are telling people who actually own/will own this stuff how they should use it? You don't see me telling you how to run your stuff?

 

[nRollo]

Old Fart:

Looks pretty darn slow in UT 2K3 in these [ H ] screenshots. 1/2 to 1/3 the FPS of a 9800P. Whats the deal?

First of all, that's just one benchmark on one site, which does not "proof" make. Second, who would really dispute the 9700/9800 series aren't top shelf cards that do some things better than the competition? I loved my 9700 Pro, kicked serious butt, I truthfully told the guy I sold it to that I considered it the best VGA I ever had, by far.

As far as UT 4X/8X goes:

At 4x/8X Anand found no discernible performance difference

Tom found the same, cards about equal

HardOCP, guess what? About the same here too

You cannot compare 6XAA to 8X AA, AA has a too great a performance hit. You can't compare 6X, because nVidia doesn't have it, and you can't compare 8X because ATI doesn't have it. So you're left with 4X for direct compare.
I notice it doesn't show tell us the resolution of your "important" screenshots. As that's UT, and at 12X10 4X both the 9800 is at 92fps on one map, and 73 on another, we can assume this is at 10X7 or below?

When I buy a 5900Ultra, I plan to try to run at 12X10, not 10X7?

 

[oldfart]

First of all, that's just one benchmark on one site

No, it was not a benchmark. This is actual gameplay at both cards "max quality" setting. Yeah, they dont support the same level of AA and AF. ATi's 6X AA looks better than nVidia's 8x AA. To be fair, the ATi AA should have been LOWERED to 4x to match the lower AA quality of the 5900U to really make a valid comparison. The test should not have an equal "AA x value". They should have an equal "AA quality level".

I dont know what the res was. I wish they would have included that iinfo.

 

[kylebisme]

if you want to know the res of a screenshot, rightclick on it and relect properties.

 

[Keysplayr]

I think I have had it with the forum scene. Time to move on. I just sat back last night and read and read. I thought, "my god, everyone on here is
ridiculous, including me. My god, im turning into a loser geek with nothing better to do.". I have no life so I sit late at night and pick fights or whatever.
How lame. Anyway, this will be my last post in any forum. Don't even reply to this post because I will never see it. I am gone from you guys. escaping if you will. So, keep on doin whatever it is your doin. Just wanted to say, Goodbye.

Keysplayr2003...................................... fini

 

[nemesismk2]

Originally posted by: keysplayr2003
I think I have had it with the forum scene. Time to move on. I just sat back last night and read and read. I thought, "my god, everyone on here is
ridiculous, including me. My god, im turning into a loser geek with nothing better to do.". I have no life so I sit late at night and pick fights or whatever.
How lame. Anyway, this will be my last post in any forum. Don't even reply to this post because I will never see it. I am gone from you guys. escaping if you will. So, keep on doin whatever it is your doin. Just wanted to say, Goodbye.

Keysplayr2003...................................... fini

Goodbye, you have my sympathy. If I didn't have to visit forums like this for my job I wouldn't be here either!

 

[nRollo]

f you want to know the res of a screenshot, rightclick on it and relect properties.

Thanks Snowman!
Old Fart: So it is at a lowly 10X7. Answer me this Old Fart: Is it better in terms of image quality to run at low resolution, really high AA, or high resolution, low AA?