Here's what AMD didn't want us to see - HAWX 2 benchmark

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
To be exact, it was AMD's mail asking people NOT to use the benchmark, which triggered the HardOCP post which started all this: http://hardocp.com/news/2010/10/20/benchmark_wars
Nvidia asked reviewers TO USE the benchmark, which they co-wrote. And they literally dropped off overclocked 460's at reviewers doorsteps at the last minute. Not to mention the "review guides" Nvidia is infamous for. Don't act like Nvidia is some innocent entity. They wrote the book on dishonest practices, and I suspect as they bleed more and more market share, they will take things to even more dishonest levels.
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
Nvidia asked reviewers TO USE the benchmark, which they co-wrote. And they literally dropped off overclocked 460's at reviewers doorsteps at the last minute. Not to mention the "review guides" Nvidia is infamous for. Don't act like Nvidia is some innocent entity. They wrote the book on dishonest practices, and I suspect as they bleed more and more market share, they will take things to even more dishonest levels.


AMD has review guides too...you know that right?

And it's not like AMD didn't push DX11 tesselation benchmarks when NVIDIA had no hardware..AvP or DIRT2 anyone?
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Folks, a plea for some civility if you will.

If you make it personal or take it personal then odds are that you are not enjoying yourself here, and we want you to be here because you enjoy being here.

Please keep the content of your posts on-topic and not inflammatory or flame-baiting. Seriously guys, you are making baby Jesus cry.

Moderator Idontcare
 

Scali

Banned
Dec 3, 2004
2,495
0
0
You should read the link you provided. It clearly states, "Yesterday we got an email from NVIDIA about a new H.A.W.X. 2 benchmark and the company encouraged us to use it on upcoming reviews."

Yea I read that.
Then I read the next line:
"Today we got an email from AMD asking us not to use it on upcoming products."
And apparently THEN they posted this article, as I said.

I think you should read the link as well.
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
I think you should read the link as well.
Would you find it acceptable if AMD assisted a game dev to make a demo that purposely used a code path that crippled performance on Nvidia hardware, and then asked reviewers to include it as part of their benchmark suite? Also, any offer Nvidia made to the dev to help improve performance would be denied.

All good?
 

Scali

Banned
Dec 3, 2004
2,495
0
0
Well, let's put it this way:
If you are a hardware company, what is more suspicious?
When you offer people hardware and point them to new benchmarks because you WANT them to benchmark your products against the competition?
Or when you send reviewers emails saying that you DON'T want them to compare your products against the competition?
 

Scali

Banned
Dec 3, 2004
2,495
0
0
Would you find it acceptable if AMD assisted a game dev to make a demo that purposely used a code path that crippled performance on Nvidia hardware, and then asked reviewers to include it as part of their benchmark suite? Also, any offer Nvidia made to the dev to help improve performance would be denied.

All good?

Okay, I see that in your head it is already decided that nVidia has deliberately crippled performance on AMD's hardware.
Apparently in your head there is no possibility that AMD's hardware might actually perform considerably worse regarding certain aspects of the DX11 standard.

As long as you are not doing anything nasty, I have no problems with making benchmarks and games that show off the latest features on new hardware.
I fully supported ATi with their Radeon 9700, and 3DMark03, Half-Life 2 and various other games, which pushed SM2.0, making GeForce FX look like the underpowered piece of you-know-what that it really was.
Nothing wrong with that. As soon as nVidia came out with the GeForce 6-series, their SM2.0 was up to snuff, and all those benchmarks and games worked fine as well, because there was nothing nasty about them, no crippling nVidia.
 
Last edited:

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Would you find it acceptable if AMD assisted a game dev to make a demo that purposely used a code path that crippled performance on Nvidia hardware, and then asked reviewers to include it as part of their benchmark suite? Also, any offer Nvidia made to the dev to help improve performance would be denied.

All good?

I would find it acceptable if AMD or Nvidia did that. (I'd be worried about their competitive spirit if they didn't try)

What I would be worried about knowing, as a consumer, is whether or not my trusted review site(s) complied.

You guys are familiar with the story of the parable The Scorpion and the Frog?

The story is about a scorpion asking a frog to carry him across a river.

The frog is afraid of being stung, but the scorpion reassures him that if it stung, the frog would sink and the scorpion would drown as well.

The frog then agrees; nevertheless, in mid-river, the scorpion stings him, dooming them both.

When asked why, the scorpion explains, "I'm a scorpion; it's my nature."

It is foolish to expect either business entity to be anything but the scorpions they aim to be.

At the same time I expect my trusted reviewers to let me know when something is afoot so I am not the stupid-ass frog of the story!
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
Okay, I see that in your head it is already decided that nVidia has deliberately crippled performance on AMD's hardware.
I'm going by pattern of behaviour. Nvidia has in the past (Batman AA etc.) indeed intentionally caused lower performance on AMD hardware. In the case of Batman AA, ATI cards actually do all the rendering passes, but never benefit from the output. Using the hardware ID hack, performance stays the same, but anti-aliasing is performed identically to an Nvidia card.

Nvidia has set the precedent, they can't be trusted.
 

Scali

Banned
Dec 3, 2004
2,495
0
0
I'm going by pattern of behaviour. Nvidia has in the past (Batman AA etc.) indeed intentionally caused lower performance on AMD hardware. In the case of Batman AA, ATI cards actually do all the rendering passes, but never benefit from the output. Using the hardware ID hack, performance stays the same, but anti-aliasing is performed identically to an Nvidia card.

I have already commented on this numerous times, even having to argue with people who don't even understand what deferred shading is.
Short version: you're wrong.
 

Lonbjerg

Diamond Member
Dec 6, 2009
4,419
0
0
I have already commented on this numerous times, even having to argue with people who don't even understand what deferred shading is.
Short version: you're wrong.


I have noticed at trend.
AMD PR makes a claim.
The claims gets debunked.
AMD fans still use original (now debunked) claims as a "fact".
Makes you wonder...
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
I would find it acceptable if AMD or Nvidia did that.
I personally don't find it acceptable at all, especially from the game dev.
I'd be worried about their competitive spirit if they didn't try
That's one way to look at it. Or, a company may just be confident enough in their hardware that they don't need to resort to tricks and cheats. Ultimately, the market will decide what is correct.
At the same time I expect my trusted reviewers to let me know when something is afoot so I am not the stupid-ass frog of the story!
Me too. Which is why I am disappointed that Ryan Smith used an overclocked 460, even after stating that it explicitly goes against review policy.
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
I have already commented on this numerous times, even having to argue with people who don't even understand what deferred shading is.
Short version: you're wrong.
I've tested the scenario myself. You are wrong.
 

Meghan54

Lifer
Oct 18, 2009
11,684
5,222
136
Cool!

Weak geek fights! Or more like a kitten/bitch slap fight.

Thanks for the entertainment.....and all the angst about a video card, something that doesn't matter one whit in life.

Makes you wonder who shills for whom seeing who is taking what side in the petty arguments above.


Thread-crapping is not acceptable.

Moderator Idontcare
 
Last edited by a moderator:

Scali

Banned
Dec 3, 2004
2,495
0
0
I've tested the scenario myself. You are wrong.

So you tried a hack and gave it a look, and figured "Hum, okay, I think this looks like it's supposed to be looking"?
Do you even know how the hack works, and why it works?
Do you know what it is you're supposed to be testing in the first place?
 

GaiaHunter

Diamond Member
Jul 13, 2008
3,650
218
106
So you tried a hack and gave it a look, and figured "Hum, okay, I think this looks like it's supposed to be looking"?
Do you even know how the hack works, and why it works?
Do you know what it is you're supposed to be testing in the first place?

Yes.

Since the Game of the year edition is out you can see the hack allows exactly the same AA quality.
 

SirPauly

Diamond Member
Apr 28, 2009
5,187
1
0
It is foolish to expect either business entity to be anything but the scorpions they aim to be.

I agree!

Personally expect both do be disingenuous and it is not what they do say but what they do not say. Personally expect both to shine lights on what they do well and cast shadows on their competitors and shine lights back on themselves. The idea of good guy vs bad guy are really more-so illusions and they're corporate entities, and what is best for the company's self interest.
 

Scali

Banned
Dec 3, 2004
2,495
0
0
Yes.

Since the Game of the year edition is out you can see the hack allows exactly the same AA quality.

I already said it was trivial for the developer to make it work on DX10.1 hardware in another thread.
You didn't answer my questions though:
What does the hack do, and how do you test if it works?
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
So you tried a hack and gave it a look, and figured "Hum, okay, I think this looks like it's supposed to be looking"?
No. I compared the identical save point/location between Nvidia and ATI cards. One with the hack, one without. Without, the ATI card had no anti-aliasing. With the hack, it did, and the screenshot was identical to the Nvidia card. The frame rate on the ATI hardware remained identical with or without the hack. The only difference is AA was applied vs. not.

I don't know if I still have the screenshots, I doubt it. But I'm not the only one that did these tests, it's easily repeatable.
 

Scali

Banned
Dec 3, 2004
2,495
0
0
No. I compared the identical save point/location between Nvidia and ATI cards. One with the hack, one without. Without, the ATI card had no anti-aliasing. With the hack, it did, and the screenshot was identical to the Nvidia card. The frame rate on the ATI hardware remained identical with or without the hack. The only difference is AA was applied vs. not.

I don't know if I still have the screenshots, I doubt it. But I'm not the only one that did these tests, it's easily repeatable.

You don't understand.
I'm asking you if you understand why the hack would work.
Because once you understand why the hack works, the rest is easy to explain. As long as it doesn't, you'll remain stuck in your fantasy world that this is all nVidia's evil-doing.
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,979
589
126
You don't understand.
I'm asking you if you understand why the hack would work.
Because once you understand why the hack works, the rest is easy to explain. As long as it doesn't, you'll remain stuck in your fantasy world that this is all nVidia's evil-doing.
I know exactly why the hack works. The vendor ID check is nullified/moot, so the rendering path no longer jumps over a stage, and the proper output is realized.