nVidia should nix 3dmark

[Matthyahuw]

I think nVidia should write a code to detect 3dmark and not allow it to run.

I think Futuremark is getting too big, and I think we need to shut them down.

That will stop all the cheating on both sides dead!

We'd then have to use *GASP* games to benchmark with!

[volt]

3DMark is a joke. Burn in hell and die!

Thank you :cool:

[Hanners]

Originally posted by Matthyahuw
We'd then have to use *GASP* games to benchmark with!

Then IHVs would start cheating in game timedemos (If they aren't already).

[Matthyahuw]

Can't cheat in all games all the time, that's why you bench more than two games...

or then start using FRAPS...

[Hanners]

Ok, let's say we do that....

How are we going to test Pixel Shader 2.0 performance? And performance of other DirectX 9 features?

[Matthyahuw]

why do we need to?

wait for a game to actually USE them!

Until then, it's moot...

watch the company demos and go OoOoOo AHHHHHH

[Hanners]

Some people can't afford to upgrade to every new generation of graphics cards, they'll buy one card to last them several years - In that kind of scenario, future performance is very important.

That's pretty much the whole reason nVidia want to discredit 3DMark2003 in my opinion - To hide the absolutely appaling Pixel Shader 2.0 performance of their current range of cards.

[evilangel]

I also think Nvidia should not use 3DMark. We know HardOCP doesn't like it, if more sites just stopped using it, FutureMark goes down the toilet.

The people who make up the majority of video card sales don't care about Pixel Shader 2.0 performance or 3DMark.

[Sazar]

Originally posted by evilangel
I also think Nvidia should not use 3DMark. We know HardOCP doesn't like it, if more sites just stopped using it, FutureMark goes down the toilet.

The people who make up the majority of video card sales don't care about Pixel Shader 2.0 performance or 3DMark.

[H] still uses or at least did use elements from other parts of the test... ie... pixel and vertex shading...

the test is unlikely to just stop being used because nvidia is having problems on it @ the moment... I am sure nvidia will fix its issues and get back with the program with a better product eventually...

I had thought that the nv35 was it... but not quite it seems... perhaps the nv40 will have more and better fixes when it is released...

[digitalwanderer]

Originally posted by evilangel
I also think Nvidia should not use 3DMark. We know HardOCP doesn't like it, if more sites just stopped using it, FutureMark goes down the toilet.

Oh, well if HardOCP doesn't like it then I'd best delete it from my hard-drive right now! :eek:

:rolleyes:

Wow, you mean Kyle & nVidia are saying the EXACT SAME THING?!?! What next, what are the odds? Can we at least see nVidia's lips moving this time when Kyle talks? :p


The people who make up the majority of video card sales don't care about Pixel Shader 2.0 performance or 3DMark.

Yeah, but apparently nVidia feels it's important enough to cheat at to improve their score. :rolleyes: :lol:

[evilangel]

I don't trust FutureMark, I don't take their word on anything including the cheating. They're probably just pissed off at Nvidia because they're down $200,000.

Bottom line the bulk of the market doesn't care about 3DMark. It's the pseudo video card experts who sit in front of their PC either posting their pearls of wisdom that they read at another forum pretending they're a programmer. They're the minority, why cater to them? Screw 'em.



Hopefully 3DMark will be gone soon. :)

[Hanners]

Originally posted by digitalwanderer
Wow, you mean Kyle & nVidia are saying the EXACT SAME THING?!?! What next, what are the odds? Can we at least see nVidia's lips moving this time when Kyle talks? :p

Surely you aren't insinuating that nVidia have their hand up where the sun don't shine when it comes to Kyle Bennett? :lol:

[StealthHawk]

Originally posted by Matthyahuw
Can't cheat in all games all the time, that's why you bench more than two games...

or then start using FRAPS...

There are issues with those suggestions...which have been pointed out before.

Specifically, reviews already use more than two games. But review sites have a tendency to use the same games. IHVs will undoubtedly optimize for the games that big review sites benchmark.

As for FRAPS:
a) it is time consuming.
b) it produces varying results.
c) it renders comparing numbers between websites impossible.

C is particularly important, take Anandtech's NV35 preview for instance. By looking at other site's numbers, it is painfully clear that there is something wrong with his Quake3 4x FSAA numbers, as they show no decrease as resolution increases, while every other review site shows otherwise.

Discussion (http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=12343)

[hithere]

Originally posted by evilangel
I also think Nvidia should not use 3DMark. We know HardOCP doesn't like it, if more sites just stopped using it, FutureMark goes down the toilet.

The people who make up the majority of video card sales don't care about Pixel Shader 2.0 performance or 3DMark.

It's nice that you feel that way. What about those that disagree with you? Too bad, under your scenario they don't get to choose for themselves.

Is it cool to you that Nvidia goes around determining which benchmarks are used, and if one is out there they don't like, they simply discredit it by engineering cheats and hoisting thier scores on the public as valid, right up until they are finally caught, and then they make enough legal threats and hoopla that the company that made the test is discredited? As in, "if you don't like the way we do things, we'll run you out of business with lies and BS?"

That's cool that you need nvidia around to make decisions about which benchmarks should exist and which shouldn't. I don't.