• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Futuremarks Flies the NVBird:

F

fallguy

Guest
Originally posted by Kyle
Futuremarks Flies the NVBird:
Futuremark, the company behind the most widely used canned 3D benchmark on the planet, plays nice once again with ATI, but tells NVIDIA to stuff their latest version of Forceware that is WHQL and being shipped with retail cards.


ATI CATALYST DRIVERS - Latest Approved:
» ATI Catalyst 4.3 Drivers - WinXP

NVIDIA FORCEWARE DRIVERS - Latest Approved:
» NVIDIA ForceWare 52.16 Drivers - WinXP

So be warned reviewers. If you don't use "approved" drivers that give results that Futuremark wants, you will get a nasty email about your "violation." (And you might think about why you use a benchmark from a company that takes money from any IHV willing to give it to them, then let that company tell you how to benchmark.) For those of you at home, we suggest you use whatever drivers you wish.
Click to expand...

Suprised nobody has noticed this, or maybe just didnt want to post it. I guess Im a little confused. Just seems odd to me that drivers can increase the score of 3dmark2003 by 20% or more, yet does not yeild the same increase (or even close to it) in games. To me its obvious that something is going on, which is not helping games. Call it "optimizations" or whatever you want.

Seems to me that nVidia is doing whatever they need to do to get the results that they want, not Futuremark.
 
next the Nvidia FX5900 5950 will be pulling 10000 3dmarks03 but still only pulling 15 to 20 FPS in newer games:rolleyes:
Will there cheating ever end and will they ever admit that there FX range of cards are just plain horrible :confused:

Stay tooned too The nvidia Scam show lmfao:p
 
Nvidia cheats on benchmarks? I've never heard that before! :eek:




Because I've been living under a rock.




On Mars.
 
LOL you never heard of that :p

ati cheats too but ati dont show that ...
 
I agree with Kyle, use whatever drivers you like.

For my FX, I use nVidia drivers................ not Futuremark's:p
 
Originally posted by Hal|9k
Nvidia cheats on benchmarks? I've never heard that before! :eek:

Because I've been living under a rock.

On Mars.
Click to expand...
Did you get wet? :p

---------------------
Yeah, Kyle is right. Use whatever drivers you want. But the newer ForceWare drivers *do* have corresponding increases in performance in several games (TR:AoD PS2.0, Halo PS2.0, etc). Only a fanboy would say otherwise. ;)
 
Originally posted by pxc
Did you get wet? :p

---------------------
Yeah, Kyle is right. Use whatever drivers you want. But the newer ForceWare drivers *do* have corresponding increases in performance in several games (TR:AoD PS2.0, Halo PS2.0, etc). Only a fanboy would say otherwise. ;)
Click to expand...

Are these gains in a handful of games derived from compiler optimizations or shader replacements? If the latter, what about games that never become popular as benchmarks with the review sites? Will they see similar performance to other PS 2.0 games?

Moreover, how many publishers or game developers take the time to retroactively look at their released games and test each new driver drop from the IHVs to see if everything's on the up 'n up? Answer: zero, not a single one of them. Nor will they ever do so. So FM gets villified for trying to protect their program, for not being able to write code that can't be compromised (something much, much larger companies have likewise failed to do) in one way or another.

If Kyle believes that games are why we buy these expensive PC parts and that they should therefore be the focus of a solid review, I can't argue with that. [H] by far produces the best graphics card reviews amongst the major hardware sites these days. I can and always will, however, question the logic of blaming one company for the actions of another.
 
3DMark scores don't mean jack if they are used to bench different video cards. I don't pay attention to 3DMark scores at all when I am reading a review. They are useless to me because Futuremark gives the companies access to the source code. To me, that's like casinos allowing people to play with loaded dice.

The only thing I pay attension to is how the video cards do with the different games.

3DMark is useless unless you are using it to compare changes to your system or comparing systems very close to each other using the same video card.

Futuremark lost all credibility as a benchmark program when they allowed the companies access to their source code. I can't believe some sites still have no got a clue yet and still use 3DMark.

Who cares about Futurmark and 3Dmark except for testing changes on your own system. If you look at 3DMark scores in a review and go oooooo, ahhhhhh, /drool, then you need to get a clue.
 
Nvidia cheats on benchmarks? I've never heard that before!
Because I've been living under a rock.
On Mars.
Click to expand...

That means there must be liquid water there! :eek:

Moreover, how many publishers or game developers take the time to retroactively look at their released games and test each new driver drop from the IHVs to see if everything's on the up 'n up? Answer: zero, not a single one of them. Nor will they ever do so. So FM gets villified for trying to protect their program, for not being able to write code that can't be compromised (something much, much larger companies have likewise failed to do) in one way or another.
Click to expand...

Excellent points of course.
If nvidia has hand-coded shader replacements for a few key games and the IQ loss is minimal, I'd be okay with it assuming they'd let you choose, but allowing such a choice would paint their hardware in an unfavorable light.

Because Futuremark has made an effort to actively defeat such tricks and manipulations in order to provide a level playing field, they get trashed at every opportunity because no one else even bothers to make an effort. :rolleyes:
 
here is the question. if they are trying to defend their software, than why are they giving the source out to anyone who can foot the bill, it seems to be the exact opposite of what they should be doing.
 
Who says they are giving out the source code? Even if they are, how does it make it any different? Do you think nvidia would not cheat without it. If so, that still makes futuremark the ones in the wrong and not nvidia?
 
Futuremark allowing access to the benchmark source code is like a teacher allowing access to the test he/she is going to give.

BTW, ATI cheated also so when you mention Nvidia, also remember to mention ATI. They both guilty.
 
Originally posted by Fetti
Futuremark allowing access to the benchmark source code is like a teacher allowing access to the test he/she is going to give.
Click to expand...

Except in this instance the student rapes the teacher. It's like Soviet Russia.
 
Originally posted by Fetti
Futuremark allowing access to the benchmark source code is like a teacher allowing access to the test he/she is going to give.
Click to expand...

It's nothing like that at all. Give me a real world example. Nvidia bailed out of the beta program but still cheated their asses off anyways. You're full of it. Your excuses for nvidia make no sense at all.

Originally posted by Fetti BTW, ATI cheated also so when you mention Nvidia, also remember to mention ATI. They both guilty. [/B]
Click to expand...

Well that debateable in the first place and was removed from the very next driver revision. I was talking about the present, nvidia still cheat to this day and it doesn’t look like they will ever stop.
 
While I don think 3dmark2003 is a "Gamer benchmark" like they say it is. Mainly because it gives a higher score to a 1gig Celeron and a 9800XT than a 3gig P4 and a 9700. The 3gig+9700 is going to be faster every time in games, yet the benchmark says its not.

It DID give some early results of PS 2.0 and the FX's being well below that of the R3x GPU's. And we find out today that it is true.
 
Originally posted by fallguy
While I don think 3dmark2003 is a "Gamer benchmark" like they say it is. Mainly because it gives a higher score to a 1gig Celeron and a 9800XT than a 3gig P4 and a 9700. The 3gig+9700 is going to be faster every time in games, yet the benchmark says its not.

It DID give some early results of PS 2.0 and the FX's being well below that of the R3x GPU's. And we find out today that it is true.
Click to expand...

Well fair enough with it not really being a gamers benchmark. It is more of a straight up video card benchmark. But I remember 2001 being the same thing, back when it first came out and all our systems were slow. Only the geforce 3 could do it all back then. that was the top of line just out. Once we are on our nextgen video cards and all past 10,000 marks it will start to turn into what 3dmark 2001 is for us now or more like it was a year or so ago.

Anyways, this thread is more about putting the blame on futuremark for graphics companies cheating in their test. It seems absurd, I really can't see how it's their fault.
 
yes, but 2k1 scaled not only with vid cards but also with processors and ram, in 2k3 it doesn't matter if you have a 400 mhz p2 or an athlon fx, same vid card will get about the same score.
 
Anyways, this thread is more about putting the blame on futuremark for graphics companies cheating in their test. It seems absurd, I really can't see how it's their fault.
Click to expand...

Precisely. One IHV is unrepentantly manipulating the benchmark using disallowed program-specific optimizations and somehow that's the developers fault.

Remember people, global driver optimizations that impact other programs and do not degrade image quality are allowed by futuremark.

But nvidia isn't doing that, so complaining that newer drivers also increase performance on unrelated games means nothing.

Either these optimizations are not global driver optimizations and do not spill over to 3DMark, or they degrade image quality.
It's quite likely nvidia has resorted to hand-coding shader substitutions once again for 3DMark, meaning they've pulled engineering resources away from general driver optimization specifically to cheat.

I cannot see how that could be viewed as a good thing by anyone, regardless of whether it increases their numerical score in a tired and rather contrived video card benchmark.
 
Originally posted by Fetti
Futuremark allowing access to the benchmark source code is like a teacher allowing access to the test he/she is going to give.

BTW, ATI cheated also so when you mention Nvidia, also remember to mention ATI. They both guilty.
Click to expand...

BTW remember ATI admitted to it and took it right out which was a 1%-2% drop then, nothing major. When Nvidia removed it, it was a huge drop in some test. And ever since Nvidia has been putting these cheats back in and out because without them thier cards are horrible with DX9 shaders. I'm shady of the 56.56 under all the games like Kyle said Tomb Raider, Halo. All of a sudden the card is right on par with the radeon when before the radeon jumped ahead. I'm sure its more cheats or "optimizations", basically taking code and changed it to thier benefit.

DaveM
 
Originally posted by Fetti

3DMark is useless unless you are using it to compare changes to your system or comparing systems very close to each other using the same video card.

Who cares about Futurmark and 3Dmark except for testing changes on your own system. If you look at 3DMark scores in a review and go oooooo, ahhhhhh, /drool, then you need to get a clue.
Click to expand...
Amen.

One other thing, why did HardOCP condem nVidia last year for cheating, but now make it seem like it's ok? Is it because they are more pissed at FutureMark than nVidia now? Kinda jumping both sides of the fence aren't we?

3DMark cannot be ignored by OEM's and hardware manufacturers. Just because we know it's a invalid benchmark, there are scores of people and websites and magazines that publish results from it. And even more people do than don't make purchasing decisions based on it's results. I think FutureMark is justified in making jugements on which drivers get used in official tests.

It would be kind of nice to back a few years and check and see if nVidias 20% performance increase on the GF3 was actual driver improvements or cheats like they are now....
 
I'll buy whoever's card is better damn it. And that includes IQ, not just #'s. /rubs a magic lamp and makes it all go away.
 
You must log in or register to reply here.
Back
Top