Cheating or Not, who is and who isn't? ATI? Nvidia?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

SilverTrine

Senior member
May 27, 2003
312
0
0
Whats this who cares crap? Of course I care if the card I buy cheats in major benchmarks. That would be like Intel acting like a 2.0ghz processor performs like a 2.4ghz processor by cheating in benchmarks, better believe I care.
As of now the ONLY drivers that detect 3dmark2003 and run custom code are Nvidia's.
So theres your answer Nvidia is the only company still engaging in cheating.
 

Keysplayr

Elite Member
Jan 16, 2003
21,219
55
91
Lock this thread!!!!! Its been so done. Why cant you just do a search for the other 250 threads that deal with this madness? There is no solution and if you think AT forum can give you one, you are barking up the wrong IP address.
 

SneakyStuff

Diamond Member
Jan 13, 2004
4,294
0
76
This thread has absolutely NO RELEVANCE TO ANY REAL-WORLD COMPUTING TASKS. I just play the games, I dont give a flying fvck who "cheats" with their drivers, as long as they're compatable, and they run my games, i'm content. And like nick said, they both cheated, nobody really cares, end of story! :D
 

Ackmed

Diamond Member
Oct 1, 2003
8,499
560
126
You cant just say, "They both cheated, the end." nVidia still does, their drivers detect 3DMARK, which is a violation of FM's rules. nVidia knows it, FM knows it, only the ignorant reviewers seem to not know it. They still post results with 53.03 drivers, which falsely inflate the score of FX cards.

Case in point: http://www.driverheaven.net/reviews/XTs/index.htm

The Point Of View 5900XT comes clocked at 300/680

They are not even smart enough to know the card has 2D and 3D clocks, that are different. They have edited their review now to show the correct speeds. I mean honestly, where do they get these guys?

They even went on to say that the 53.03 drivers were not FM approved, yet used them for 3DMARK 2003 anyway. They have since taken the 3DMARk 2003 portion out of their review. Guess they got embarrassed. :(

 

JJN

Member
Dec 28, 2003
48
0
0
optimize this, render that, blah blah blah the cards aren't exactly the same and therefore things don't look exactly the same. Some cards are better at certain aspects than others. Just put more stock in the reviews that use actual gameplay for their benchmarks and stop relying on crap like 3DMark03.
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
I think we never got anywhere with these arguements because of people that say "who the hell cares" that distract us from the real problem.

3dmark03 is just a bunch of bs, them an their fake gaming benchmark that couldn't be more from their advertised truth. It should be a call a GPU benchmark that tests the power of GPUs. 3dmark2001 deserves more merit. Still used and more of the kind of benchmarks, FM claims 03 to be.

Editing in-game screenshots is also another form of cheating or a really persistant bug. But, I doubt it.
 

Keysplayr

Elite Member
Jan 16, 2003
21,219
55
91
Everyone form a circle and close your eyes and pray. Maybe the Gods of AT will answer our prayers. Lock, lock, lock... Chant with me....
 

cm123

Senior member
Jul 3, 2003
489
2
76
Originally posted by: VIAN
I think we never got anywhere with these arguements because of people that say "who the hell cares" that distract us from the real problem.

3dmark03 is just a bunch of bs, them an their fake gaming benchmark that couldn't be more from their advertised truth. It should be a call a GPU benchmark that tests the power of GPUs. 3dmark2001 deserves more merit. Still used and more of the kind of benchmarks, FM claims 03 to be.

Editing in-game screenshots is also another form of cheating or a really persistant bug. But, I doubt it.


Why can't there be a bench mark using DirectX that fails unless ALL requirements of the ver. of directX are being used during the test.

I like MS Rise of Nations, it shows the frame as you play, a fair way to see the performance as used I feel.

Must be a way to do the same for openGL

 

BFG10K

Lifer
Aug 14, 2000
22,709
3,003
126
The definition of a cheat is quite simple:

A section of code that relies on pre-calculated knowledge (e.g algorithmic detection, application detection, shader detection, etc) to work.

and/or

A section of code that for all possible inputs purposefully alters one or more outputs to incorrect values.

Anyone who violates those rules is cheating, regardless of who it is.
 

Ackmed

Diamond Member
Oct 1, 2003
8,499
560
126
All of your "lock this thread" posts, just bump it up to the top. Good job on that.
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
There is no God.

The definition of a cheat is quite simple:

A section of code that relies on pre-calculated knowledge (e.g algorithmic detection, application detection, shader detection, etc) to work.

and/or

A section of code that for all possible inputs purposefully alters one or more outputs to incorrect values.

Anyone who violates those rules is cheating, regardless of who it is.

I agree with those. Application detection should not be required. Even if it is to fix a problem, the developer should provide a fix and not the Hardware manufacturer.

and others call that cheating if there drivers to whatever (including lowering grahpics quality), why is that a cheat?


It's not, but some people seem to think it is. Although, I wouldn't like the graphics card to do that automatically. The good it performs or how crappy it looks should be up to the user. If I want real Trilinear, then I want real Trilinear, not Brilinear or Bilinear where Trilinear should be. It is a cheat if I set the filter to Trilinear and they give me bilinear. That is breaking the rule I just set as well as an attempt to change the definition of Trilinear.
This is what I would like to request from Nvidia and ATI. I think the list is complete here.
 

nick1985

Lifer
Dec 29, 2002
27,153
6
81
Originally posted by: Ackmed
All of your "lock this thread" posts, just bump it up to the top. Good job on that.

actually, i wouldnt have posted that if it hadnt already been at the top. so i really didnt bump it did i? you dolt?
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
actually, i wouldnt have posted that if it hadnt already been at the top. so i really didnt bump it did i? you dolt?
Still makes a difference dolt.

You give us many posts and makes the thread seem more important. Thanks.
 

BFG10K

Lifer
Aug 14, 2000
22,709
3,003
126
I agree with those
Then you'd have to class nVidia as cheating.

Application detection should not be required.
I think it's acceptable to resolve compatibility issues. Some games just won't behave under standard rendering paths and in the end the consumer loses out as a result of this. Application detection in such a case benefits the consumer and because it's not motivated to simply inflate benchmark results, it's a good thing.
 

nick1985

Lifer
Dec 29, 2002
27,153
6
81
Originally posted by: VIAN
actually, i wouldnt have posted that if it hadnt already been at the top. so i really didnt bump it did i? you dolt?
Still makes a difference dolt.

You give us many posts and makes the thread seem more important. Thanks.

your welcome VIAN! thanks for using good manners. i now value your otherwise worthless opinions.


BTW, your mothers a whore.
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
What do you think about sending this email to both ATI and Nvidia:

This is what myself and some gamers at a forum that have purchased your hardware think is cheating and should be changed if you are guilty. We also include some ideas that we feel would better your product. This is not directed toward any specific company.

1. A section of code that relies on pre-calculated knowledge (ex.: algorithmic detection, application detection, shader detection, etc) to work.

2. A section of code that for all possible inputs purposefully alters one or more outputs to incorrect values.

Number two includes such things as using "Brilinear" instead of Trilinear. If the application calls for a certain setting and/or we choose a certain setting, then we expect to get that setting and not a cheap imitation. We feel cheated by those cheap imitations. Instead, if you wish to continue with the optimization, may you please give use the option to turn it off as well as giving a name to your optimization and not using a name that already means something. An example of this is Nvidia's Trilinear, dubbed "Brilinear." If it isn't the same, then why is it called the same name. Give us the option to have the game rendered just as the application states. Hardware limitations may be taken into consideration for using alternate values.

Please take these statements into consideration. We hope we will be seeing these changes in the very near future. Thank you for time.

your welcome VIAN! thanks for using good manners. i now value your otherwise worthless opinions.
I stopped caring about your posts a while ago. They don't mean anything and are just wasting good bytes.

BTW, your mothers a whore.
Thank You.

V isually
I mpaired
A nal
N ewbie
I love you. :lips:
 

cm123

Senior member
Jul 3, 2003
489
2
76
Originally posted by: VIAN
What do you think about sending this email to both ATI and Nvidia:

This is what myself and some gamers at a forum that have purchased your hardware think is cheating. We also include some ideas that we feel would better
your product. This is not directed toward a specific company.

1. A section of code that relies on pre-calculated knowledge (ex.: algorithmic detection, application detection, shader detection, etc) to work.

2. A section of code that for all possible inputs purposefully alters one or more outputs to incorrect values. This includes such things as using
"Brilinear" instead of Trilinear. If the application calls for a certain setting and/or we choose a certain setting, then we expect to get that setting
and not a cheap imitation. We feel cheated by those cheap imitations. Instead, if you wish to continue with the optimization, may you please give use
the option to turn it off as well as giving a name to your optimization and not using a name that already means something. An example of this is Nvidia's
Trilinear, dubbed "Brilinear." If it isn't the same, then why is it called the same name. Give us the option to have the game rendered just as the
application states. Hardware limitations may be taken into consideration for using alternate values.



__________________________________

Not a bad idea...
 

VIAN

Diamond Member
Aug 22, 2003
6,575
1
0
feel free to write any additional changes that you would like to include or take out.
 

SneakyStuff

Diamond Member
Jan 13, 2004
4,294
0
76
OMFG!!!!!!!!!! nobody cares! if YOU want to send a letter to them super, but COME ON! That is the pinnacle of geekdom........
rolleye.gif