3DMurk03 new cheats ?

[mczak]

I think this needs some more investigation. Remember, for instance it makes a difference if you choose aniso in 3dmark03 or force it in the driver (at least for ATI cards it's faster if it's forced in the driver). This is strange, and so far I haven't seen an explanation why this is the case, though it's possible 3dmark03 uses different mip-filters if you enable aniso (afaik it doesn't use trilinear for everything without aniso, but it might if you switch it on). That's however pure speculation, and I'd like to know what is really going on.
So I'd like to see some benchmark numbers and image differences between aniso 3dmark03 / aniso driver, both with 3dmark and 3dmurk (and also with an ATI card, I'd just like to know what's the difference if aniso is enabled in the driver or in 3dmark03 itself).

 

[micron]

We already have dialouges with these people so I feel confident we can get things done.

"We"?

He has a turd in his pocket....

LOL! That made my morning.
Thanks!

Your welcome!

 

[Miksu]

Anyone tried renaming some other benchmark/game to 3DMark03.exe and running it? It should do something to the game, shouldn't it.

 

[Humus]

I think it very perculiar that tech report tested only the patched version of 3dmark2003 - ie: this could be a trap laid for nVidia by futuremark, with the help of their sponsors ATi.

While I don't think there's any truth in this I'm not going to bash you for it like some other people in here. In fact, it's not that far-fetched, though I would have worded it differently. The source was ATI, and it's nice to know that the source was presented and it's good that ET left the conclusion upon the reader. Since the source is a competitor there's obviously more interests involved than just the truth. Checking that the source is clean too is important. A few tests with the unpatched version should be enough. Though I'm certain there wouldn't be any conspiracy revieled by doing so.

 

[epicstruggle]

Anyone tried renaming some other benchmark/game to 3DMark03.exe and running it? It should do something to the game, shouldn't it.

now that would be interesting. i would imagine they check the size of the exe.

later,

 

[Captain Chickenpants]

Anyone tried renaming some other benchmark/game to 3DMark03.exe and running it? It should do something to the game, shouldn't it.

Interesting point, although it would rely on them using just the appname, which would mean the change would affect all the tests. I suspect it would be more likely to check the app name + some other combination of settings.

But if it is just the app name, then it should be fairly simple for someone to knock up an app to highlight the changes.

CC

 

[breez]

Anyone tried renaming some other benchmark/game to 3DMark03.exe and running it? It should do something to the game, shouldn't it.

now that would be interesting. i would imagine they check the size of the exe.

later,

Checking the size is not the case here because tests were done on build 330. Unless the 320 and 330 exe size is same (which I highly doubt).

 

[RussSchultz]

What I'm a little disappointed in is that Futuremark didn't catch this in their original audit. Its the simplest check to do.

These are the same drivers available before the original ET article, no?

 

[micron]

What I'm a little disappointed in is that Futuremark didn't catch this in their original audit. Its the simplest check to do.

These are the same drivers available before the original ET article, no?

If I was going to be disapointed in anybody, it wouldnt be Futuremark :?

 

[Nazgul]

I wonder what the FutureMark guys are thinking right now, with this new information coming to light so soon after they caved in to Nvidia's strong-armed tactics(yeah yeah, we don't -really- know what happened, but c'mon...).

And Kyle's response over at [H] is truly pathetic. If Nvidia pulling something like this further invalidates 3DMark2003, then why didn't he pull Quake3 out of his benchmark suite after ATI and the whole "Quak" fiasco? I find it amazing how much he's just not getting who the bad guy in all this really is.

 

[micron]

I have tried to reason with Kyle, and question him on his ethics.
I questioned his reasoning enough times that he banned me. POOF!
I have come to the conclusion that Nvidia has made him feel like a special boy somehow, and he really liked it. He will defend them untill it is no longer possible. If Nvidia is in fact publicly proven to be cheating in games and timedemo's, Kyle will take a stand.

 

[Doomtrooper]

If you don't like it, don't go there...Like Terry Markedon stated..ATI PR decides who is getting review cards soley off traffic vs. content. I couldn't believe it but it's true.

Tomshardware Hits

Today 1 wk. Avg. 3 mos. Avg. 3 mos. Change 
      1,346    1,281     1,394       136

Anandtech Hits

Today 1 wk. Avg. 3 mos. Avg. 3 mos. Change 
      2,341    1,940     1,924       141

Hardocp Hits

Today 1 wk. Avg. 3 mos. Avg. 3 mos. Change 
      3,352    3,743     3,874       1,713

What website has been on the Editorial rampage with ANTI-3Dmark campaigns..then look at the numbers above in the last three months..

All information is from Alexa..

http://www.alexa.com/data/details/traffic_details?q=&url=http://www.hardocp.com/

 

[RussSchultz]

Heh. You guys should stop visiting so often to check on things to bitch about.

 

[radar1200gs]

Anyone tried renaming some other benchmark/game to 3DMark03.exe and running it? It should do something to the game, shouldn't it.

Interesting point, although it would rely on them using just the appname, which would mean the change would affect all the tests. I suspect it would be more likely to check the app name + some other combination of settings.

But if it is just the app name, then it should be fairly simple for someone to knock up an app to highlight the changes.

CC

A good idea.

The anisotropic tester utility used by several web sites including this one should make a very good candidate application for renaming.

The allegation is that nVidia's drivers alter anisotropic filtering when they detect 3DMark03.exe - what better way to test this?

 

[Joe DeFuria]

What I'm a little disappointed in is that Futuremark didn't catch this in their original audit. Its the simplest check to do.

Yes and no...

It requires that you run the tests with Aniso on (and perhaps only at 8X setting?). It's understandable that they could miss this.

If scores changed WITHOUT Aniso activated, that would be really disappointing.

Heh. You guys should stop visiting so often to check on things to bitch about.

Yup. I won't even visit their forums....

 

[micron]

I think Alexa is that little registry bug that my Ad-Aware keeps finding.
I pick one up every time I go to one of those bukake pr0n~ sites

 

[jandar]

I think Alexa is that little registry bug that my Ad-Aware keeps finding.
I pick one up every time I go to one of those bukake pr0n~ sites

So you've been studying nVidia's new secret blur vision?

 

[micron]

I think Alexa is that little registry bug that my Ad-Aware keeps finding.
I pick one up every time I go to one of those bukake pr0n~ sites

So you've been studying nVidia's new secret blur vision?

Yes, they call it Nblur.

 

[YeuEmMaiMai]

Here are the results with ATi's cat 3.4 drivers

The following settings were the same for both runs

2xFSAA
2xANSIO
Texture Preference Quality
Mipmap level Highquality
Vsync OFF
TruForm OFF

Radeon 9500 Pro clock speeds 360/306

3dmark score 2856
3dmurk score 2854

A whole 2 point difference or .07% drop

Tasks running in background

Fold@Home (dos box)
Promoxitron
Rage3d Tweak
Outlook
IE (3 seperate instances)

__________________
ECS K76SA, Win 2K SP2, 512MB DDR2700 ram, AMD Athlon XP 1.7G clock, Radeon 9500Pro, 120GB WDC Caviar 7200rpm drive, SB Audigy Platnum, USR 56K PCI hardware modem, Lite-On 40/12/48 CD-RW, Toshiba 12xDVD- Reigon free, ViewSonic A90F

Busted! (again?)

Looks like ATI have decided that have had enough of being the 'mr. nice guy' and have taken the gloves off, so to speak.

Completely out of sense of fairness, has anyone tried this with ATI drivers? You gotta figure though if ATI released this info, they'd have made pretty damn sure that their driver don't do it too!!

Funny though, I don't see this story splashed all over the front pages of [H]. Wouldn't this be a, errm, optimisation that could be just as easily applied to all benchmarking-capable games?

 

[radar1200gs]

Go here:
http://briefcase.yahoo.com/gregrstanford
(you will need a yahoo account if you don't already have one).

Go to the Anisotropic folder and download "Container.zip"

It will have 2 ace files and two jpg files in it.

The two jpg's show my driver settings, used for the test.

Aniso Tester.ace is Xmas's anisotropic filtering testapp V1.2 as used by this website.

Aniso1.ace contains 2 tiff images.

The first of these is the anisotester run normally with the settings indicated in the picture.

The second is the anisotester renamed as 3DMark03.exe (which is the exact filename 3dmark 2003 installs as on my system).

Note I only have a GF3 Ti200 currently.

I'd like to see results from GF-FX owners please.

Remember the allegation is that nVidia's drivers detect the 3dmark executable name and then modify their anisotropic filtering settings based on the name detected. If this in fact true, this test will pick it up.