Motherboards.org uses 53.03 benching 3dm2k3?!?

[Crash_7]

(long time lurker, first time poster)

Probably a stupid question, but wouldn't it be possible for Futuremark to set up 3dmark2003 so that it simply wouldn't run with unapproved drivers? You know, have it call home to Futuremark and check the drivers on the system running the benchmark against a database of currently approved drivers. It might be a bit inconvenient for review sites, but it's a lot better than FM having to chase down every bogus benchmark in every review.

 

[Tim Murray]

(long time lurker, first time poster)

Probably a stupid question, but wouldn't it be possible for Futuremark to set up 3dmark2003 so that it simply wouldn't run with unapproved drivers? You know, have it call home to Futuremark and check the drivers on the system running the benchmark against a database of currently approved drivers. It might be a bit inconvenient for review sites, but it's a lot better than FM having to chase down every bogus benchmark in every review.

Of course it's possible, but that seems a little draconian. One of the things about review machines is that they often don't have active Internet connections.

 

[Quitch]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

 

[Tim Murray]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

And the original poster's idea relies upon 3DMark phoning home every time it's run, not simply patches.

 

[digitalwanderer]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

And the original poster's idea relies upon 3DMark phoning home every time it's run, not simply patches.

Or you could just have it say "UNOFFICIAL" in big red flashing letters on the results screen until such a time as it was posted to the ORB, and when it's posted to the ORB it'll become official and the "UNOFFICIAL" in big red flashing letters won't show up anymore.

Simple & effective.

 

[Tim Murray]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

And the original poster's idea relies upon 3DMark phoning home every time it's run, not simply patches.

Or you could just have it say "UNOFFICIAL" in big red flashing letters on the results screen until such a time as it was posted to the ORB, and when it's posted to the ORB it'll become official and the "UNOFFICIAL" in big red flashing letters won't show up anymore.

Simple & effective.

That's the obvious thing to do, but reviewers do not always rely upon the ORB.

 

[micb]

Hello worm,

Here is another site benching with the 53.03.


http://www.rojakpot.com/showarticle.aspx?artno=74&pgno=0


It is on the page, 3 - Test setup & Benchmark settings page.

Graphics Processor NVIDIA GeForce FX Go5200
Graphics Driver NVIDIA Detonator 3 Ver. 53.03

 

[ByteMe]

]

Yo Worm! Do you think that maybe Futuremark isn't doing enough to get the word out about which drivers are and are not acceptable when beta members of Futuremark don't even know 'em?

We are doing the best we can to get the word out on what drivers are to be used in reviews with 3DMark03. I have no idea why this particular review has used drivers which are not approved by us. :? I'll look into this asap.

Why bother? That is like trying to patch a hole in a damn with your thumb. Why not fix the problem instead of the result?

 

[Quitch]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

And the original poster's idea relies upon 3DMark phoning home every time it's run, not simply patches.

Well, no, it would only need to update its "approved" driver file, and if your driver wasn't approved you could download the latest file to see if it was now. It would hardly take an age, could be put on the CD with the driver, and at worst would produce a warning box which I'm sure there could be an option to disable.

But at least then people have to go out of their way to ignore it, not out of their way to find it.

 

[Joe DeFuria]

Why bother? That is like trying to patch a hole in a damn with your thumb. Why not fix the problem instead of the result?

Because the problem is nVidia. And FM doesn't control them.

 

[Nick[FM]]

Or you could just have it say "UNOFFICIAL" in big red flashing letters on the results screen until such a time as it was posted to the ORB, and when it's posted to the ORB it'll become official and the "UNOFFICIAL" in big red flashing letters won't show up anymore.

Simple & effective.

We are updating the ORB (and all official lists) at the moment to point out which results are got with approved drivers, and which aren't. It is indeed sad that so few reviewers submit the 3DMark03 results to the ORB, and provide compare URL's in their reviews. If they would do that, it would be much easier for everyone to see all numbers, and if the result is valid or not.

 

[dan2097]

Cant you just sue Nvidia for loss of business or something?, thats getting to the root of the problem.

 

[MuFu]

Oh, and post 1,000. Hercules! Hercules!

Gigabyte! Gigabyte!

<groan> <groan>

Are you playing with yourself?

MuFu.

 

[digitalwanderer]

What kind of excuse is that? We live in an age of USB Pens, CD Burners, Ethernet connections. They manage to shift the latest driver onto the review machine, but the latest version of the software they're using to test is too much trouble? Don't make me laugh.

And the original poster's idea relies upon 3DMark phoning home every time it's run, not simply patches.

Or you could just have it say "UNOFFICIAL" in big red flashing letters on the results screen until such a time as it was posted to the ORB, and when it's posted to the ORB it'll become official and the "UNOFFICIAL" in big red flashing letters won't show up anymore.

Simple & effective.

That's the obvious thing to do, but reviewers do not always rely upon the ORB.

True, they usually just grab screenshots of their final scores anymore to "prove" it's official...thus the "UNOFFICIAL" watermark on the final score tally screen until it's posted to the ORB.

I know it ain't perfect, but it would help at least remind reviewers...they'd KNOW they're doing an endrun around FM by just charting the results at least.

]

Or you could just have it say "UNOFFICIAL" in big red flashing letters on the results screen until such a time as it was posted to the ORB, and when it's posted to the ORB it'll become official and the "UNOFFICIAL" in big red flashing letters won't show up anymore.

Simple & effective.

We are updating the ORB (and all official lists) at the moment to point out which results are got with approved drivers, and which aren't. It is indeed sad that so few reviewers submit the 3DMark03 results to the ORB, and provide compare URL's in their reviews. If they would do that, it would be much easier for everyone to see all numbers, and if the result is valid or not.

I still think you'd be better of just not accepting any un-approved drivers scores to the ORB, that would pressure nVidia in a big way by having all the fanboys cry about not being able to post their "big" scores with their nV hardware. When they backlash against FM over not being able to point 'em at the real problem and explain to them to complain to nVidia.

 

[cthellis42]

Are you playing with yourself?

Um... Ew?

 

[Crash_7]

Of course it's possible, but that seems a little draconian. One of the things about review machines is that they often don't have active Internet connections.

That's what I meant about it being inconvenient for review sites. But it just seems to me Futuremark needs to do something about this if they want to remain a viable benchmark. It's obvious from the number of reviews that have been pointed out using unapproved drivers with 3dmark2003 that the concept of having an approved driver list simply isn't working. Futuremark can go scurrying all over the web trying to correct bogus benchmark results, but once they've been posted in a review, the damage has already been done even if it is corrected later.

 

[ByteMe]

The next time I get sick I must remember to kick the dog.

 

[Tim Murray]

How about this--no results are displayed immediately. They are either uploaded to the ORB and verified or they are saved to an encrypted file on the drive, which can then be taken to a computer with an Internet connection, uploaded, and verified? Sure, it'd be cracked eventually, but I don't think you're going to get De3DMark from your favorite Belgian teenager anytime soon.

But anyway, with that solution, the only way you can see results is with Futuremark's seal of approval.

 

[Althornin]

ARG!
[H] misses the point, again!

Futuremark Wants Benchmarks Removed:
OverclockerCafe has a short and to the point editorial posted on a situation that they experienced today with Futuremark concerning 3DMark2003 results. It seems as though drivers that are released to the public are not good enough to benchmark with. Or do the drivers "cheat" at 3DMark03 scores? If you ask us, we will simply tell you that 3DMark03 is showing its true worth.

We started with our campaign for better benchmarks well over a year ago and have written an editorial or two along the way. What we have come to find is that canned benchmarks are worthless to HardOCP and that we were simply doing the wrong thing "reviewing video cards." We finally figured out that we need to be evaluating the gameplay those videocards supply. Our 3D Editor, Brent Justice, has picked up the reigns and been producing solid content that we feel is worthwhile to more people.

Too bad your content ignores the future performance of said reviewed products. Synthetic benchmarks ahve proven thier worth again and again, and i hold [H] responsible for the huge amount of nonsense floating around about them. A valid predictive tool is being thrown away for NO FREAKING REASON.



OverClocker Cafe's editorial:
http://www.overclockercafe.com/Futuremark.htm

 

[kyleb]

from overclockerscafe:

It almost comes across as if Futuremark is happy only as long as ATi isn’t besting NVidia on its benchmark.

the moron thinks it looks like futuremark is standing up against nvidia's "optimizations" because they don't want ati winning the benchmarks? how can he even think that makes sense?!?