"Quack3"? nVidia Now Brings "3DMurk" To Customers

[BFG10K]

Oh man, this just keeps getting better and better. Hot on the heels of the proven cheating in 3DMark (most notably the reduced pipeline precision, static clip planes and shader subsitution) comes the new "3DMurk" cheat.

Basically renaming the 3DMark executable and running anisotropic filtering with it causes reduced scores as nVidia is reducing the level of anisotropy when it detects 3DMark is running with anisotropic filtering.

Oh dear.

 

[zephyrprime]

A pox on them. This all reminds me of that crazy 286 cheating scandal long, long ago.

 

[nortexoid]

and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw.

u have to wonder whose side those scoundrals are really on - the humans' or the mutants'?

 

[Czar]

and the fight goes on

 

[ndee]

hot damn now here we go again

 

[Czar]

saw this on their comments section

Secondly Id also find ways of testing Q3, Serious Sam and Unreal Tournement. If you'll notice Nvidia had Big gains is perormance in AA+AF in those games to. But strangely enough only in Certian levels. The ones the Big sites use. When you look at some of the minor review sites that use different levels for Testing they cards are Tied or the 9800pro beats the 5900. the 9800pro also wins the Splinter cell and Warcraft benchmarks which are not widely used either.

any comments?

 

[PrinceXizor]

Hmm...one wonders if renaming your benchmarking program will now become standard procedure for reviewing cards. Sad really.

P-X

 

[bgeh]

that's sad, first ATi, now nVIDIA

 

[Whitedog]

Originally posted by: nortexoid
and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw.

u have to wonder whose side those scoundrals are really on - the humans' or the mutants'?

This is Pentium of Borg... Precision is Futile! You will be approximated.

 

[Double Trouble]

I don't know about the rest of the folks out there, but I'm getting really tired of the companies trying to cheat to gain an edge in the bechmarks. I don't care about one company or another, I'm not a zealot saying one company is better than another, I just want the best performance for my $$$.

The companies keep finding ways to try and cheat to make their products look better in benchmark scores. Instead of wasting resources on that, why not spend the resources on actually making a better product that simply *is* better??

Press release: newly formed conglomerate nViATI announces the newest product, the Benchie 2005 XLT -- it gets the greatest scores in the history of benchmarking, it cruises through futuremark benchmarks..... Only one slight problem, it's a terrible performer in real use, it just functions when benchmarking.

Anyway, to me, when a company resorts to using gimmicks to try and improve their performance in some bencmark, it's basically saying "hey, our product can't stand up to the competition on it's own, so we have to try and create these cheats". Who'd want to buy products from that company??

 

[Harabecw]

both products deliver great performance, IMO. Its just the race to be the #1, hence being the new brand name to get. nVidia just wants to keep that by getting on top of the Radeons.

 

[spongebobfan]

I think that nVidia purposely fiddled with their drivers to invalidate 3dMark as a viable benchmark. They are trying to focus on real game benches and breaking 3dMark was a requirement. It seems they were willing to accept the short-term PR blowback (remember the ATI "cheats"; everyone remembers them but ATI is still selling more than ever). I had to laugh in the joint press release when it was said that nvidia used no "legal" pressure to get futuremark to take out "cheat". It is plainly obvious that nvidia did use "market" pressure by threatening to make 3dMark irrelevent for reliable benchmarking uses and create simply a fancy tech-demo for graphics cards. This driver "cheat" would then have the desired effect of scaring futuremark to the bargaining table where nVidia would play peacemaker and release a "reliable" driver after having made their point... Rambling....

My .02

Sponge

 

[ElFenix]

Originally posted by: Czar
saw this on their comments section

Secondly Id also find ways of testing Q3, Serious Sam and Unreal Tournement. If you'll notice Nvidia had Big gains is perormance in AA+AF in those games to. But strangely enough only in Certian levels. The ones the Big sites use. When you look at some of the minor review sites that use different levels for Testing they cards are Tied or the 9800pro beats the 5900. the 9800pro also wins the Splinter cell and Warcraft benchmarks which are not widely used either.

any comments?

see, i said in the other thread that if they were cheating in 3dmark it wouldn't be much of a leap for them to be cheating on other benches and it would be hard to detect. i then advocated the use of, if not different engines or games, then something other than the standard benchmark on every site.

 

[pm]

and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw.

No it didn't. The original Pentium had a faulty entry in the floating-point division look up table that caused it to return an incorrect value. After the issue was discovered, Intel, with some reluctance, offered to replace the faulty parts with corrected ones. There was no performance difference between the faulty and correct parts. On one it would return one value and on the correct one it would return the correct value.

 

[squidman]

WHERE MY NFORCE DRIVER AT???!

 

[Tab]

Now we will never really know which card is actually faster! AHHH!

 

[WarCon]

Sure you will, you just have to buy both........

 

[sandorski]

Originally posted by: Whitedog

Originally posted by: nortexoid
and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw.

u have to wonder whose side those scoundrals are really on - the humans' or the mutants'?

This is Pentium of Borg... Precision is Futile! You will be approximated.

ROFL

 

[AgaBoogaBoo]

Originally posted by: sandorski

Originally posted by: Whitedog

Originally posted by: nortexoid
and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw.

u have to wonder whose side those scoundrals are really on - the humans' or the mutants'?

This is Pentium of Borg... Precision is Futile! You will be approximated.

ROFL

ROFLLMAO

You should get a :beer: for that one!

 

[BFG10K]

any comments?

Yes, I've long suspected that the same cheats used in 3DMark by nVidia are also being used in the popular game benchmarks. The static clip planes in particular are more than likely to be used elsewhere.

and the original pentiums had this flaw with certain numerical operations which may have caused it to perform faster than it would've without the flaw

It's nothing like that at all. Some original Pentiums (75 MHz and 90 MHz IIRC) had bugs in the FPU which sometimes caused inaccurate values to be returned after a certain position after the decimal place (three or four, I forget the details). Also the performance was unaffected.

Think about it, if Intel had done something like that on purpose and just to increase performance it would have totally ruined them. You just don't go around making faulty CPUs to boost performance if you plan on staying in the CPU business.

 

[TheBoyBlunder]

hah. *gets out marshmallows* this should be a good ol' fashioned flame war.

 

[GTaudiophile]

So, the year-old R3XX core still wins the day???

 

[nortexoid]

guys, guys....i was joking about the fpu issue causing increased performance.

in fact, i'm never not joking...well, usually almost always sometimes.

 

[mikable]

Originally posted by: Whitedog

This is Pentium of Borg... Precision is Futile! You will be approximated.[/quote]


Funniest thing I've seen in a long, long time! makes the thread worth reading!!!