[AMD_Robert] Concerning the AOTS image quality controversy

[sirmo]

What about the part where you imply NVIDIA is cheating without having any proof? Oh snap

Which part is that?

 

[LTC8K6]

Here's what I want to know, how did Nvidia know to fix the issue if it was press only?

Did the press report the bug to Nvidia and not to the public? Or did they find the bug on their own,maybe they already knew?

Well, I thought immediately to question the driver, since a later numbered driver didn't seem to support pascal.

So I thought the driver had to be different, or perhaps it didn't support Pascal either?

If the thought occurred to me, it surely occurred to NV, and the media/reviewers?

 

[sontin]

.19 was the press driver for reviewers. The .25 is the first public driver for Pascal.

 

[swilli89]

What about the part where you imply NVIDIA is cheating without having any proof? Oh snap

Maybe their historical "driver optimizations"? Remember the cheating on 3dmark 2006? The geforce 6 series getting caught red handed reverting to reduced rendering quality levels to pump up benchmark results. I saved one entire summer in highschool to afford a 6800gt, few weeks in my games look VASTLY different than my friends with AMD X800 cards. Sure enough, read a few articles that their drivers had hidden "optimizations" (nv speak for cheats) in their consumer drivers. Went ATI/AMD never looked back...

 

[LTC8K6]

.19 was the press driver for reviewers. The .25 is the first public driver for Pascal.

Yes, we know.

 

[dogen1]

Maybe their historical "driver optimizations"? Remember the cheating on 3dmark 2006? The geforce 6 series getting caught red handed reverting to reduced rendering quality levels to pump up benchmark results. I saved one entire summer in highschool to afford a 6800gt, few weeks in my games look VASTLY different than my friends with AMD X800 cards. Sure enough, read a few articles that their drivers had hidden "optimizations" (nv speak for cheats) in their consumer drivers. Went ATI/AMD never looked back...

Both companies drivers perform lots of optimizations. Image quality differences have been found on both sides.

 

[BlitzWulf]

Both companies drivers perform lots of optimizations. Image quality differences have been found on both sides.

Yes and every time it's caught the company in question is rightly lambasted in public you cant use something AMD may or may not have done in the past to deflect criticism from something that involves Nvidia today,

This thread is about Visual and possible perf differences in AOTS with NV press release drivers not a companies do this all the time so what's your gripe? thread

 

[dogen1]

Yes and every time it's caught the company in question is rightly lambasted in public you cant use something AMD may or may not have done in the past to deflect criticism from something that involves Nvidia today,

This thread is about Visual and possible perf differences in AOTS with NV press release drivers not a companies do this all the time so what's your gripe? thread

And I say if there's no performance difference, there's no issue. No consumer is going to use the press driver.

 

[BlitzWulf]

And I say if there's no performance difference, there's no issue. No consumer is going to use the press driver.

Benchmarks that people use to make purchasing decisions are done on press drivers.

I'll agree with you on that no perf no foul.

Now who's gonna be the brave one in the press to test it out and report it if it's a result which doesn't paint a favorable picture of Nvidia?

 

[poofyhairguy]

And I say if there's no performance difference, there's no issue.

I think everyone can agree on that. The disagreement will come if there is a performance difference, and what that means.

 

[sontin]

Looking at the nVidia picture i would force nVidia to give me the cheat option (back). There is no reason to do necessary work for nothing.

 

[BlitzWulf]

Looking at the nVidia picture i would force nVidia to give me the cheat option (back). There is no reason to do necessary work for nothing.

Deflection and apologia at it's best, your subjective visual preference is not important.

The GTX 1080's performance being compared to other GPU's that were doing more work is the topic

 

[Minkoff]

I'm sure AMD started testing the moment they got GTX 1080 in their hands, as I'm certain Nvidia does/will do as well with Polaris, Vega etc.

The first available driver is the press driver. By th time of the RX490 presentation (or sneak peak) probably hundred (or more) benches have been run, are you gonna re-run them all?

Resolutions, Point Light Quality, Glare Quality, Shading Samples, Terrain Shading Samples, Shadow Quality, Multi-sample Anti-Aliasing...

Whether or not, the press driver has significant differences with the official one, we'll soon see. I kinda doubt, it, though...

...stuff..Also why did AMD use a driver that they were not even supposed to have?..stuff...

I doubt the drivers are under NDA, especially after the GPU has been released.

 

[sontin]

Sure it is. When a developer uses the most inefficient way to render terrain (why do they need to use multple shaders instead of a texture?!) then i would trade it off for more performance. It is funny that Oxide doesnt give me this option. But i guess they are trying very hard to do so much work in every frame to put AMD at the number one spot.

And yes: Even as a Maxwell user i want this option. Snow or no snow - it doesnt matter because graphics is not important here.

 

[swilli89]

Looking at the nVidia picture i would force nVidia to give me the cheat option (back). There is no reason to do necessary work for nothing.

Unbelievable. People split absolutely minute hairs on this forum over differing types of advanced AA, yet then we have a massive rendering error in a cutting edge modern DX12 title and you are basically saying "no big deal!"

 

[swilli89]

Sure it is. When a developer uses the most inefficient way to render terrain (why do they need to use multple shaders instead of a texture?!) then i would trade it off for more performance. It is funny that Oxide doesnt give me this option. But i guess they are trying very hard to do so much work in every frame to put AMD at the number one spot.

And yes: Even as a Maxwell user i want this option. Snow or no snow - it doesnt matter because graphics is not important here.

Ok wow.

 

[boozzer]

Unbelievable. People split absolutely minute hairs on this forum over differing types of advanced AA, yet then we have a massive rendering error in a cutting edge modern DX12 title and you are basically saying "no big deal!"

ahaha pretty incredible right? right.

 

[nurturedhate]

The thought process that some people are trying to use in this thread is ridiculous. Stuff like this being allowed to even exist damages the credibility of the forum and main site.

 

[poofyhairguy]

Whether or not, the press driver has significant differences with the official one, we'll soon see. I kinda doubt, it, though...

Me too, that would be a huge scandal.

 

[dogen1]

Sure it is. When a developer uses the most inefficient way to render terrain (why do they need to use multple shaders instead of a texture?!)

Don't be silly.

It is funny that Oxide doesnt give me this option... And yes: Even as a Maxwell user i want this option.

Agreed

Unbelievable. People split absolutely minute hairs on this forum over differing types of advanced AA, yet then we have a massive rendering error in a cutting edge modern DX12 title and you are basically saying "no big deal!"

It's a pretty minor rendering error if you ask me. Not that it doesn't matter of course.

 

[swilli89]

The thought process that some people are trying to use in this thread is ridiculous. Stuff like this being allowed to even exist damages the credibility of the forum and main site.

So much this. I wouldn't be any less harsh on AMD for stooping to that. I will say that there is no proof of malicious intent on nvidia's part but that's simply fuel for a forum argument, closer to the real world I think we all know what was going on here. Nvidia is getting slaughtered in AOTS, that we know is a fact. AOTS MIGHT offer a preview to DX12 performance which certainly makes it more important than if it were just another random DX11 title.

So knowing this, nVidia definitely focuses on this game's performance. You either have to claim either A) Nvidia done just goofed it up! Total screw up and not intentional, or B) Nvidia's vast army of software engineers knew exactly what they were doing and planned for this. I think if we are being fair and honest it seems argument B is more likely as nVidia is a very competent firm, they plan and execute like no one else in the industry right now so I think they would have easily caught this one in the development pipeline as you can LITERALLY see the difference in screenshots/video.

 

[Minkoff]

...stuff... But i guess they are trying very hard to do so much work in every frame to put AMD at the number one spot.

And yes: Even as a Maxwell user i want this option. Snow or no snow - it doesnt matter because graphics is not important here.

I truly apologize for the OT
So, options like...to reduce the over tessellated Batman cape, or hidden tessellated water. Oh, wait the Witcher 3 hair, no no...the concrete slabs in Crysis...
Yeah, I want those options as well..

 

[sontin]

Unbelievable. People split absolutely minute hairs on this forum over differing types of advanced AA, yet then we have a massive rendering error in a cutting edge modern DX12 title and you are basically saying "no big deal!"

"massive rendering error"? Have you looked at the pictures? The snow shader is destroying information which has been rendered by the GPU in a previous step.

The snow shader has no positiv impact on the picture. Applying a snow shader after the rendering of the terrain has no benefit at all and cost only performance.

So yes, it is no big deal because the picture looks (much) and performs better.

I truly apologize for the OT
So, options like...to reduce the over tessellated Batman cape, or hidden tessellated water. Oh, wait the Witcher 3 hair, no no...the concrete slabs in Crysis...
Yeah, I want those options as well..

You have those options in your driver. And sometimes AMD is taking care of it for you.

 

[Krteq]

.19 was the press driver for reviewers. The .25 is the first public driver for Pascal.

Wait a second, are you trying to say nVidia cheated DX12 results in AotS in all launch reviews with this .19 driver?

 

[Bacon1]

You have those options in your driver. And sometimes AMD is taking care of it for you.

And yet you call that cheating, even though the IQ is the same. Yet here Nvidia is removing image quality and you are praising them.