New 3dmark03 patch - nVIDIA cheating ... again???

[Blastman]

So I gather by all this ... Nvidia = cosmic singularity.

 

[Rogodin2]

"As for the Universal statements, the way I see it a statement is true as long as it is validated in one instance."

Wrong, read some philosophy of science.

Read some Popper and Fenyman.

Any layman can prove a theory "false" or "true" in one instance (scientific observation has to be rigidly proven true in all empirical and theoritical propositions to be valid) the one (astrology), but the real litmus test is to prove the proposition (observation) TRUE in every test (example, test, observable data).

I'm not even going to ask you about how one believes that what one observers (choice) is circumscribed by (truth).

The "truth" of nvidia's cheating has been proven, the data is valid and it is observable-this is just another premise to support the conclusion.

This is logic.

rogo

 

[Rogodin2]

My philosophy is an aglomeration of many thinkers-even though you don't "give a rat's ass" about philosophy you are inundated with "life" and this demands a philosophy wether you like it or not.

If I may suggest a wonderful and easy to read philosopher I would suggest "Jaspers" or "Keirkregaard".

rogo

 

[Blastman]

Originally posted by: NYHoustonman
ooo lotsa big vocabulary methinks peeps need to admit that what NVidia and/or ATI has been doing is wrong and it doesn't matter whether or not 3DMark matters because all that matters is that NVidia is intentionally skewing things to get more buyers like the Apple G5 and that isn't a very nice thing to do but for some reason people find a need to defend it by either saying it doesn't matter cos everything looks the same or that 3DMark doesn't matter but the fact is without "cheats" it should be able to test raw GPU speed rather than ability of driver teams to cut corners, and to the average gamer 3DMark scores may mean alot more and here NVidia is essentially lying about the speed of their graphics cards to get profit which I guess makes sense but it's not a nice thing to do and steroids aren't allowed in baseball so you see the use of analogy here because that is giving an unfair advantage just like here and maybe if all cards were forced to render the same exact thing every time we would have a better judge of what is more powerful but because of all this 3DMark now means nothing just like alot of other benchmarks where there are potential cheats and if neither company used cheats everything would be better but the world isn't perfect but it should be and i guarantee theres philosphical meaning to all this


It's amazing how far this argument goes each and every time somebody brings up such a subject... Am I the only one who couldn't give a rat's ass about philosphy?

Man, that must be the longest run-on sentence I've ever seen. LOL.

 

[jiffylube1024]

Ugh - this thread has disintegrated into the ugliest dialectic imaginable, and totally off topic of course.

 

[Rogodin2]

Yep, run-ons are the curse of intellectual "observations" on public forums; we know the intricacies of this human.

It's the scourge of modern public schooling-these poor children can's spell or write correctly, grammar is an unknown utility to these poor kids.

I see it all the time on public forums.

rogo

The dialectic consists of the thesis-antithesis-and synthesis.

The term doesn't apply to this thread AT ALL; if it did this thread would be worth reading.

rogo

 

[DefRef]

Nice to see all the college boys showing off what they can regurgitate from their classes. Very "Good Will Hunting" if ya axe me.

I've been laying off here lately because the ATI goons couldn't stop crying like little girls whenever I called them "Fanchimps", but in a thread with 80+ posts, I find it interesting that not a single person noticed the first question that sprang to my mind:

IS FUTUREMARK SABOTAGING NVIDIA BY ADDING CODE THAT WORKS LIKE "IF DRIVER=NV THEN FUX0R THEM!"?

Nvidia's been under the microscope by it's haters and the feedback on the 52.16 drivers has shown that they've upped performance without trashing IQ. Things were quieting down and SUDDENLY (dun-dun-DONG!), FM releases a "patch" that seems to do one thing only: It trashes Nvidia's 3DM score.

How convenient.:disgust:

I know the 'chimps feel sad and blue if a day doesn't go by without them shrieking that Nvidia's cheating, but their blind hatred of Nvidia has robbed them of any rational ability to ask legitimate questions. Why is it that when no one seems to have any serious gripes about the 52.16 drivers IQ and speed (other than FX cards are still not as fast) BEFORE FM released their "patch" that when Nvidia's scores falter, it's automatically assumed that FM stopped NV from cheating when it's possible that FM is out to gank NV? Has ATI sent some checks to FM lately like they cut for Valve to slant the code their way?

If anything, it appears that ATI isn't satisfied that they have the better products at the moment and seek to damage Nvidia reputation by paying off various interests to code against Nvidia. Shouldn't you be concerned or are your biases preventing you from opening your eyes?

The first step of troubleshooting an issue is to find our what's changed since it last worked. The 52.16s worked fine in other apps and now they're slow in the one app that's just changed it's code and changed the rules to prevent old code from being used. Doesn't honest inquisitiveness demand that FM explain what they did or must Nvidia attempt to prove a negative and prove they weren't cheating?

Don't talkf about "honesty" if you're only going to demand it from one player in a three-player game.

 

[spam]

Hi Insomniak,

You do like to stir the pot don't you? If you want to discuss existentialist philosophy perhaps you could visit the Off Topic forum, that may be a better venue for this conversation. There must be some pragmatic approach to evaluating the value of a given video card. That is the purpose of this forum.

 

[Rogodin2]

Defref

Are you an idiot?

your post is certainly a validation of this attiribute.

Just because you have over 2000 posts on this crappy forum does it really mean that you can call people "chimps" and fanboys without repercussions?

I think not, I'll continue to call you out on your shoddy claims anytime I see your spurious conclusions and immature propositions.

I'm not even going to quote your turgid post.

I'm just disgusted with the quality of this halfasssed forum.

If you just want to trash a person (or persons, your universal circumscription is completely irrational) use the PM, I won't because I've had it with "oldtimers" that are 12 years old and that feel and believe have the right to crap all over any thread they feel justified. Unless I get banned I will continue to call out all of you irrational ones.

Your stooping to "fanchimps" is the most conspicious symbol of your age, your maturatity, your philosphy, and the size of your brain.

I don't usually use ad hominem but your pompous 12 year old opinion deserves it.

rogo

 

[DefRef]

Yep, it goes to show that people have no use for the truth if it's not candy-coated and fed to them in accordance with their own prejudices. Poor baby. Someone better call the waaaaaaaaaahbulance for him!:disgust:

While he was ignoring my point while obsessing on my lack of interest in coddling the ATI fanatics, I was running the new 3.40 3DM2K3 and it's clear that something's up.

Test_____v. 3.2_____v. 3.4_____Reduction in %

3DMark Score_____5660 _____4457_____21.25%

CPU Score_____670_____598_____10.75%

Game Tests
GT1 - Wings of Fury_____179.5_____170.3_____5.13%

GT2 - Battle of Proxycon_____41.6_____31_____25.48%

GT3 - Troll's Lair_____33_____25.9_____21.52%

GT4 - Mother Nature_____32.4_____21.9_____32.41%


CPU Tests
CPU Test 1_____72.6_____66.9_____7.85%

CPU Test 2_____12.3_____10.6_____13.82%


Feature Tests
Fill Rate (Single-Texturing)_____1457.6_____1381.2_____5.24%

Fill Rate (Multi-Texturing)_____3140.7_____2942.3_____6.32%

Vertex Shader_____18.7 _____11.8_____36.90%

Pixel Shader 2.0_____47_____44.3_____5.74%

Ragtroll_____20_____16.5_____17.50%


Sound Tests
No sounds_____40.3_____35.8_____11.17%

24 sounds_____36.3_____35.8_____1.38%

60 sounds_____33.7_____32.9_____2.37%

(Both tests run on a non-Ultra 5900 w/52.16 clocked at 420/850 and the CPU at 193x11.)

Questions:

1. If the patch is supposed to only FUBAR GPU "optimizations", why are the CPU tests docked an average 10%?

2. What exactly are these optimizations being "corrected" by FM? Why won't they come out and say exactly what was being done improperly - i.e. like the clipping plane deal before? If the images are being properly rendered, then why does FM care so much how it's getting done unless they have an interest in hobbling one of the players?

3. As mentioned before by someone else, why is it that the same driver goes from being a "cheater" to being "approved" with only a patch to 3DM2K3? That would be like a race car having an illegal restrictor plate and the race track laying down sticky tar in that car's lane instead of demanding the plate be changed to be in compliance. WTF?

4. In nearly three weeks, no one has reported anything dire in the 52.16 driver with regards to IQ cheats, yet a simple "patch" has taken performance to an all-time low - my rig running at default CPU speed on the old 45.23 drivers posted a 5203 score without any OCing, a 14% improvement over the current state of affairs. Double WTF?!?!?

5. Why does the PS 2.0 score show almost the LEAST decline in speed when it goes to figure that's where Nvidia would likely cheat to make up for historical deficiencies? Hmmm? Are we to believe that Nvidia cheated on everything BUT the PS 2.0 code and how much crack are you smoking to believe it yourself?

Sorry, kiddies, but the onus is on FutureMark to explain what Nvidia was doing wrong and to prove that they aren't fux0ring their benchmark to screw Nvidia in favor of ATI. You can despise my rhetoric, but the general lack of curiousity from you Georges tells me you've already tried and convicted someone while the real killer may be getting away.

 

[chsh1ca]

Originally posted by: Insomniak

Oh I'm familiar (although I wouldn't call myself an expert by any stretch) of the leading theories pertaining to the creation of our universe. But that's just my point - they're theories, and nothing more. They aren't proven....hell, gravity is still a theory. We can't speak with authority on reality, and that is that.

Umm, yes and no. There is no "Theory of gravity". There's Newton's laws, which apply to the force of gravity here on earth, and objects trapped therein (at least, up to a certain extent), and then there is Einstein's Theories of General Relativity and Special Relativity. Newton's "Laws" have been proven wrong, and thus, should be rendered to Newton's Theories. The fact of the matter is, as our understanding and ability to view the universe grows, so does our ability to create theories based on this information. General Relativity has stood up to almost a hundred years worth of scrutiny. At higher levels, that we are now able to study, we are finding out that General Relativity doesn't explain everything, something Einstein himself understood.

All of your nonsense about whether we can speak on authority about reality aside, what does that have to do with this thread?
If you wish to send an off topic reply, please don't pollute this thread with it. Email me instead. chsh1ca@yahoo.ca.

EDIT:

Originally posted by: DefRef

1. If the patch is supposed to only FUBAR GPU "optimizations", why are the CPU tests docked an average 10%?

According to Jeff, you'll never see 3DMark report the same (or even similar) results twice. Ask him for an explanation.

PS: Jeff, I have been trying to download 3DMark03 since your first post, I reinstalled fresh recently, and wanted to see my scores using 340, but the download keeps failing for me. I will continue trying, and let you know once I've run the benchmarks.

2. What exactly are these optimizations being "corrected" by FM? Why won't they come out and say exactly what was being done improperly - i.e. like the clipping plane deal before? If the images are being properly rendered, then why does FM care so much how it's getting done unless they have an interest in hobbling one of the players?

3. As mentioned before by someone else, why is it that the same driver goes from being a "cheater" to being "approved" with only a patch to 3DM2K3? That would be like a race car having an illegal restrictor plate and the race track laying down sticky tar in that car's lane instead of demanding the plate be changed to be in compliance. WTF?

Actually, a simpler explanation is that NVidia asked them to play nice this time and not make a big deal about it (they didn't this time, unlike last), and there was likely money involved, hence the 'approved driver' bit. Integrity wise (hah, I know), Futuremark wouldn't want the egg on their face of accepting money from NVidia and to not release the patches, especially if NVidia approached them after they had started on this trend.

4. In nearly three weeks, no one has reported anything dire in the 52.16 driver with regards to IQ cheats, yet a simple "patch" has taken performance to an all-time low - my rig running at default CPU speed on the old 45.23 drivers posted a 5203 score without any OCing, a 14% improvement over the current state of affairs. Double WTF?!?!?

IQ cheats aside, I wonder if the drastic performance decrease is due to 3DM03 disabling the recompiling of the instructions somehow. I can't see that being the case, but I'm not an NVidia driver dev, so it may be possible. It would certainly explain the sudden dramatic decrease. I doubt NVidia was 'cheating' by a combined amount of about 60%. I do have to wonder how much exactly was disabled, and I agree, it would be nice if Futuremark came clean about what they found.

Sorry, kiddies, but the onus is on FutureMark to explain what Nvidia was doing wrong and to prove that they aren't fux0ring their benchmark to screw Nvidia in favor of ATI. You can despise my rhetoric, but the general lack of curiousity from you Georges tells me you've already tried and convicted someone while the real killer may be getting away.

Realistically, the onus isn't on anyone to explain anything. I would love to see an explanation, but one is not evidently forthcoming.

 

[Rogodin2]

"Yep, it goes to show that people have no use for the truth if it's not candy-coated and fed to them in accordance with their own prejudices. Poor baby. Someone better call the waaaaaaaaaahbulance for him!"

I'll take hanners and stealths findings over this crap.

This idoicy is the last straw, I'm done with this thread.

rogo

 

[jiffylube1024]

Originally posted by: Rogodin2
Yep, run-ons are the curse of intellectual "observations" on public forums; we know the intricacies of this human.

It's the scourge of modern public schooling-these poor children can's spell or write correctly, grammar is an unknown utility to these poor kids.

I see it all the time on public forums.

rogo

The dialectic consists of the thesis-antithesis-and synthesis.

The term doesn't apply to this thread AT ALL; if it did this thread would be worth reading.

rogo

Which is my point exactly.

 

[jiffylube1024]

Originally posted by: DefRef

Questions:

1. If the patch is supposed to only FUBAR GPU "optimizations", why are the CPU tests docked an average 10%?

Because the CPU test isn't a "real" CPU test - it's still a 3d scene rendered by the graphics card, it's just supposed to stress the CPU more.

2. What exactly are these optimizations being "corrected" by FM? Why won't they come out and say exactly what was being done improperly - i.e. like the clipping plane deal before? If the images are being properly rendered, then why does FM care so much how it's getting done unless they have an interest in hobbling one of the players?

No idea. I think it's a huge conspiracy against Nvidia and Futuremark is definately taking money under the table from some yet-unnamed competitor. Futuremark will gladly stake their reputation (what little is left) to make Nvidia look bad.

3. As mentioned before by someone else, why is it that the same driver goes from being a "cheater" to being "approved" with only a patch to 3DM2K3? That would be like a race car having an illegal restrictor plate and the race track laying down sticky tar in that car's lane instead of demanding the plate be changed to be in compliance. WTF?

What the heck are you saying? The same driver is approved because Futuremark disabled cheats/improper optimizations in 3dmark, that's how it goes from "cheater" to "approved".

4. In nearly three weeks, no one has reported anything dire in the 52.16 driver with regards to IQ cheats, yet a simple "patch" has taken performance to an all-time low - my rig running at default CPU speed on the old 45.23 drivers posted a 5203 score without any OCing, a 14% improvement over the current state of affairs. Double WTF?!?!?

Here's a tissue.

5. Why does the PS 2.0 score show almost the LEAST decline in speed when it goes to figure that's where Nvidia would likely cheat to make up for historical deficiencies? Hmmm? Are we to believe that Nvidia cheated on everything BUT the PS 2.0 code and how much crack are you smoking to believe it yourself?

You tell me hotshot - none of us here work in Nvidia's driver team or at Futuremark.

Sorry, kiddies, but the onus is on FutureMark to explain what Nvidia was doing wrong and to prove that they aren't fux0ring their benchmark to screw Nvidia in favor of ATI. You can despise my rhetoric, but the general lack of curiousity from you Georges tells me you've already tried and convicted someone while the real killer may be getting away.

Thanks for setting us straight again DefRef. Your thougful, non-condescending posts are always a pleasure and you always teach us so much!

 

[jasonja]

Originally posted by: zulfi

Originally posted by: Insomniak


People talk about cynicism as if it isn't the truth. Oh well. Nice guys finish last.

As for reality, accountability - that's basically a bunch of mumbojumbo.

As for improving the world that's one of the most laughable ideas I've ever heard. "Improving" things for one person or persons will only piss off another group. Welcome to Chaos Theory 101.

Look, see my other post here, the one I dragged over from Beyond 3d. That's about the size of the situation. This whole thing has been a waste of time, and thus I wash my hands of it.

The world is too dangerous to live in, not because of the people who do evil, but because of the people who sit and let it happen - Albert Einstein

Hey! you jacked my sig!

BTW, nVidia lies, cheats, pays developers millions of dollars, hurts kittens. I think it's important that Futuremark expose this. FM has a product that they wish to say is important and useful. nVidia likes to optimize for it and then show them how meaningless it is, FM has good reason to prove nVidia wrong and protect it's products credibility.

 

[spam]

Beyond 3d has posted it's results of 3dmark 340 patch patch 340
Make sure you read the conclusion where comments are posted from Nv and ATI.

B.T.W. I really like the way that this review was done. Anandtech reviewers please take note.

 

[MichaelZ]

wow, look at all the fanboys

*continues to munch on pop corn* yes yes... do go on. *makes wanking motion*

 

[JustStarting]

first hand results on a 5950 Ultra:

patch 340= 5298 pts
patch 330= 6002 pts

both performed at the same settings.

I have the only 5900/5950 using build 340- c'mon guys let me compare to something similiar!!

5950 Ultra w/ build 340

 

[JonnyBlaze]

Originally posted by: JustStarting
first hand results on a 5950 Ultra:

patch 340= 5298 pts
patch 330= 6002 pts

both performed at the same settings.

I have the only 5900/5950 using build 340- c'mon guys let me compare to something similiar!!

5950 Ultra w/ build 340

you get me in fillrate, but i smoke you in almost everything else

JBlaze

 

[JustStarting]

Originally posted by: JonnyBlaze

Originally posted by: JustStarting
first hand results on a 5950 Ultra:

patch 340= 5298 pts
patch 330= 6002 pts

both performed at the same settings.

I have the only 5900/5950 using build 340- c'mon guys let me compare to something similiar!!

5950 Ultra w/ build 340

you get me in fillrate, but i smoke you in almost everything else

JBlaze

Nice, but I was looking for a 5900/5950 comparison. I knew some fanboy would step in with a 9800 score.

 

[Tab]

I cant beileve people are arugining over this bloatware...

 

[JonnyBlaze]

Originally posted by: JustStarting

Originally posted by: JonnyBlaze

Originally posted by: JustStarting
first hand results on a 5950 Ultra:

patch 340= 5298 pts
patch 330= 6002 pts

both performed at the same settings.

I have the only 5900/5950 using build 340- c'mon guys let me compare to something similiar!!

5950 Ultra w/ build 340

you get me in fillrate, but i smoke you in almost everything else

JBlaze

Nice, but I was looking for a 5900/5950 comparison. I knew some fanboy would step in with a 9800 score.

fanboy. sure. its my first ati card in 10 years.

 

[JonnyBlaze]

Originally posted by: JustStarting

Originally posted by: JonnyBlaze

Originally posted by: JustStarting
first hand results on a 5950 Ultra:

patch 340= 5298 pts
patch 330= 6002 pts

both performed at the same settings.

I have the only 5900/5950 using build 340- c'mon guys let me compare to something similiar!!

5950 Ultra w/ build 340

you get me in fillrate, but i smoke you in almost everything else

JBlaze

Nice, but I was looking for a 5900/5950 comparison. I knew some fanboy would step in with a 9800 score.

fanboy. sure. its my first ati card in 10 years.

 

[Corporate Thug]

Originally posted by: Rage187
"And you aren't taking it seriously enough."


LOL, my video card runs and looks great. It also works w/ EVERY game out.

So, should I really be freaking out about a synthetic benchmark that does not even perform like a true dx9 game would???

No.

So why do you?

that isnt the point, it is a matter of cheating

lets say you have a A+ in a class, you are doing great, is it ok to cheat because the last test doesnt effect your grade one way or another?

 

[Insomniak]

Originally posted by: Rogodin2
"As for the Universal statements, the way I see it a statement is true as long as it is validated in one instance."

Wrong, read some philosophy of science.

Read some Popper and Fenyman.

Any layman can prove a theory "false" or "true" in one instance (scientific observation has to be rigidly proven true in all empirical and theoritical propositions to be valid) the one (astrology), but the real litmus test is to prove the proposition (observation) TRUE in every test (example, test, observable data).
rogo

Yes, any layman can. And in any identical instance with identical circumstances, the same theory will prove true again. That was my point.

As for Nvidia's cheating, that's completely on topic with this thread and off topic from our conversation. Let's shy away from it shall we? I'm not in the modd to discuss that at this point.