Does GameWorks influences AMD's Cards game performance?

[ThatBuzzkiller]

I just want to note that nearly everyone arguing that GameWorks is fine and a good thing has been proven to be an Nvidia fanboy (including you of course). However, not one of you has bothered to give a reasonable explanation of the pattern we're seeing. This kind of thing only ever happens with GameWorks titles, and we know for sure that AMD isn't allowed to optimize GW code. It only takes common sense to see what's happening here.

No reasonable explanation ?!

No, it's just you don't want to acknowledge that there ARE factors other than Gameworks such as poor developers, poor code NOT related to Gameworks, AMD's compiler, AMD's driver developer team, and etc ...

No one here is saying that Nvidia Gameworks is fine as anyone with two brain cells can clearly see the relationship, however that does not definitively mean that it is Gameworks alone causing those issues ...

 

[96Firebird]

I just want to note that nearly everyone arguing that GameWorks is fine and a good thing has been proven to be an Nvidia fanboy (including you of course).

Aww, that's cute, you call people fanboys. Run along now, let the adults talk...

 

[Techhog]

I'm the one grasping at straws yet your the one madly attesting a theory as a fact without any sort of solid grounds ?

Whatever you say ...

We have solid ground, but you choose to ignore that for the sake of your agenda. You don't have an explanation for why this pattern exists, so until you have a valid counterargument that doesn't involve some crazy coincidence, I'm going with the most logical conclusion.

 

[Techhog]

No reasonable explanation ?!

No, it's just you don't want to acknowledge that there ARE factors other than Gameworks such as poor developers, poor code NOT related to Gameworks, AMD's compiler, AMD's driver developer team, and etc ...

No one here is saying that Nvidia Gameworks is fine as anyone with two brain cells can clearly see the relationship, however that does not definitively mean that it is Gameworks alone causing those issues ...

So, you're saying that there's no relation? What are the odds of that? Okay, there's a chance that you're right. There's also a chance that we'll be attacked by aliens who have been spying on us using cows for centuries.

Aww, that's cute, you call people fanboys. Run along now, let the adults talk...

You cut out the rest of that post, so I guess that you have no counterargument. I'll take that as your surrender.

I've doubled my ignore list today...

Are you finished? Now add this thread to your ignore list, and get out of it immediately, or you're getting banned. I'm getting sick of your crap.
-- stahlhart

 

[96Firebird]

You cut out the rest of that post, so I guess that you have no counterargument. I'll take that as your surrender.

I've doubled my ignore list today...

Once I read the first part, I didn't have to read the rest. I don't debate with people who can't help but label others.

 

[Magee_MC]

It would be much more believable that the hidden GW code didn't actively degrade the performance of AMD GPUs if NV hadn't been caught doing exactly that in the past.

http://www.vrworld.com/2009/11/04/batmangate-amd-vs-nvidia-vs-eidos-fight-analyzed/

With the description above being explained to us as "standard", it turns out that the code that was labeled "MSAA Trademark NVIDIA Corporation" was identical to the one recommended by both sides with one very important difference: nVidia’s code features a Vendor ID lock that allows for AA menu inside the game only if nVidia hardware is detected.

What got AMD seriously aggravated was the fact that the first step of this code is done on all AMD hardware: "’Amusingly’, it turns out that the first step is done for all hardware (even ours) whether AA is enabled or not! So it turns out that NVidia’s code for adding support for AA is running on our hardware all the time – even though we’re not being allowed to run the resolve code!
So? They’ve not just tied a very ordinary implementation of AA to their h/w, but they’ve done it in a way which ends up slowing our hardware down (because we’re forced to write useless depth values to alpha most of the time&#8230!"

 

[stahlhart]

Discuss the OP, rage about GameWorks to your heart's content -- but respect one another's opinions, even the ones you don't agree with, and stop the arguing.
-- stahlhart

 

[96Firebird]

Sorry, didn't see this post before...

Well, as it stands, the majority here seems to think so. If it isn't true, Nvidia better get their PR guys to start setting the record straight. Perception is king. True or not. Just like the perception of "poor drivers" from AMD. True or not.

I agree, both Nvidia and AMD should be more vocal when it comes to accusations like these. If Nvidia is merely providing support and features with the GameWorks program, they should reassure the public of that. If AMD is indeed getting shut out during development, they should also be vocal about that. As it is now, we have no idea what is happening, and to take one side or the other would be preposterous.

I really think these "tech journalists" should start doing some real investigation. This accusation is pretty serious. The idea of GameWorks sabotage their competition is something we, the readers/customers, should not take lightly. It is bad for gamers. Journalists really need to stand by their readers/customers and get to the bottom of this.

I also agree with this, I'm surprised we haven't seen more sites take a look at this. While he hasn't done so in an article (that I have seen), Brent Justice of HardOCP has chimed in about his take on GameWorks and it's effect on AMD...

On the topic of GameWorks, GameWork's features utilized in games and the overall impact on a game's performance between AMD and NVIDIA is over exaggerated. The usage of a GameWorks feature does not create a bias between AMD and NVIDIA hardware.

I have tested many games where GameWorks features demand the same amount of performance out of either equivalent AMD or NVIDIA GPU. There are some specific features that work more efficiently on one GPU over the other, and we point that out when we find it. The outright usage of a GameWork feature again does not create a bias between hardware. Some game's don't even use that many GameWorks features, Dying Light for example only uses 2, yes just 2, GameWorks features. These features work the same on AMD and NVIDIA GPUs since they are DX11 calls. That hardly creates bias.

AMD has every opportunity to optimize for any game they wish to invest resources in. Granted, NVIDIA's game relations and game testing and optimization labs are generally recieved as bigger with larger budgets than AMD. This is probably the most compelling reason for why NVIDIA has faster support of SLI/optimization in games than AMD does. We are constantly on both NVIDIA's and AMD's backs to optimize for new games.

In the end, our concern is for the gamer and owner of GPU hardware. Prompt support for CrossFire and SLI with good scaling and game optimizations is important for PC gamers. As a gamer, you too should be concerned about such things, afterall, you have invested your hard earned dollars into these hardware environments. You don't want your investment wasted.

Source

 

[tg2708]

What I'm reading through makes one thing clear to me is that the people defending gameworks, having no ill effects on AMD hardware cannot come up with a reasonable stance as to why it doesn't while others that say it does have a much better approach and or reasoning (providing substantiable evidence) as to why it does in many cases. I'm not loyal to any brand but it seems many of the naysayers are just defending unfair business pratices which is actually hearting consumers and just playing devil's advocate.

 

[Spanners]

Borderlands: The Pre-Sequel.

Basically an expansion pack, i.e. AMD have already sucked up their period of poorer performance penalty (PPPP&#8482, look at the Borderlands 2 release numbers.

I also agree with this, I'm surprised we haven't seen more sites take a look at this. While he hasn't done so in an article (that I have seen), Brent Justice of HardOCP has chimed in about his take on GameWorks and it's effect on AMD...

Source

TLDR:
Forums members:Repeated explanations that GW is a black box and impossible for AMD to optimize until developers play ball.

BrentJ:AMD should just optimize these games anyway.

Great back and forth, learned so much

http://arstechnica.com/gadgets/2015...thanks-to-strong-ps4-and-camera-sensor-sales/

22.3 million GCN GPUs

Maxwell, 1-2 million GPUs sold to consumers so far.

Anyone that says GameWorks harms AMD is full of it. Developers code for the largest market which is the console market. The odds are stacked against Nvidia. AMD's Sabotage Evolve program clearly hurts GeForce users as been shown many many times.

Using infantile re-phrasings of AMD's programs doesn't make your already weak argument look any better and saying "as been shown many many times" doesn't constituent it being shown at all.

 

[Silverforce11]

Forums members:Repeated explanations that GW is a black box and impossible for AMD to optimize until developers play ball.

BrentJ:AMD should just optimize these games anyway.

Great back and forth, learned so much

To be fair to BrentJ, these GW devs eventually do optimize their game and come the AMD performance patch, things are fine. It just happens a few months after release. Well beyond the typical media cycle, where the damage is done and dusted.

To gamers who don't know better, they believe the "just another game where AMD drivers bomb" mantra from NV's marketing.

Saying that, BrentJ was very harsh on AMD with Far Cry 4 lacking CF support, even after AMD went public saying they require Ubi to patch their game to enable CF to work, BrentJ kept on blaming AMD for the problem. Funny enough, once the patch was deployed and it works fine, BrentJ still bagged AMD for being late with the fix.

I just can't take someone like that seriously when they are supposedly unbiased reviewers.

 

[AtenRa]

Yes it defenitly does

I have stopped buying any GameWorks games after Watchdogs. When NVIDIA will optimize the games for its GPUs but also let the competition optimize the games for their GPUs, I will start buying those GW games again.
I dont have a problem when NVIDIA optimize the GW games for their GPUs but i do have a problem when they dont let developers optimize the game for the competition.
If NVIDIA has a new feature or architecture that needs the game to be optimized for it by all means do it, but let the competition optimize the same game for its features and or architecture. That goes for every company.

Until then NVIDIA, your GW titles and GPUs are not welcome to be considered as an alternative for PC Gaming systems. As a gamer i dont need another Intel tactics company in GPU space.

 

[R0H1T]

To be fair to BrentJ, these GW devs eventually do optimize their game and come the AMD performance patch, things are fine. It just happens a few months after release. Well beyond the typical media cycle, where the damage is done and dusted.

To gamers who don't know better, they believe the "just another game where AMD drivers bomb" mantra from NV's marketing.

Saying that, BrentJ was very harsh on AMD with Far Cry 4 lacking CF support, even after AMD went public saying they require Ubi to patch their game to enable CF to work, BrentJ kept on blaming AMD for the problem. Funny enough, once the patch was deployed and it works fine, BrentJ still bagged AMD for being late with the fix.

I just can't take someone like that seriously when they are supposedly unbiased reviewers.

Shouldn't GW titles (even some GE ones) come with a caveat that "performance may be broken, on AMD cards" before the first patch is released? I think it should be mandated by law, even for review sites!

 

[DarkKnightDude]

I thought it was obvious that gameworks affects AMD's card performance lol.

 

[BHZ-GTR]

Guys

How Game Work on the impact of AMD graphics cards . Using the code or functions

For Instance : Physix , Nvidia USE SSE Instructions .

Which part of the hardware to the limit

 

[Jaydip]

@RS I have no issues if AMD/NV makes game engines and tbh I think that would be real good.NV runs a business not a charity so why on earth they would share their codebase same with AMD? if AMD does that it would be utter stupid too.It is a war and everything is permissible here.You say about PC gaming community and all now tell me how many AAA games have you bought is the last couple of years? and why there are so many kick starters asking devs going back to their roots? its been a long time since I enjoyed so much playing POE, you want to restore PC gaming to it's root? ask the devs to create good games not just good looking games.

@3D Yes how much chest thumping we do PC gaming is irrelevant as a AAA platform.So when did GTA5 came out on consoles? BW considered the champion of PC gaming released DAI with a broken control scheme, the answer lies for you there, no one cares for PC gaming any more.

 

[Alatar]

I honestly can't take the "it must be true because some games that I've branded as gameworks games run better on Nvidia" arguments all that seriously.

Newsflash: Most of the time when gameworks is used it's tied to individual effects and settings in games that you can actually switch on and off to test the effect and difference. But instead of doing this most people will jump to conclusions.

The best example of this is watch dogs. Gameworks got a huge amount of flack for the poor AMD performance in watch dogs but most people doing the complaining conveniently "forgot" that:

1) Watch dogs only had 1 gameworks implementation that could actually be ran on AMD cards (HBAO). The rest was NV exclusive AA methods. Game had nothing else from GW in it.

2) That GW HBAO actually had pretty much exactly the same performance hit on bpth NV and AMD hardware.

3) AMD themselves released a performance driver update to watch dogs boosting performance by generous amounts (30% or something) right after launch.

But despite that the FUD had already spread and people didn't let go of it.

Another one is project CARS, it immediately got "lel gameworks doing its thing" reputation but no one will tell you which parts of the game actually feature gameworks, if those will affect AMD performance badly etc. And I personally even did testing to find out that a simple increase in the power limit will make the game get a 50% performance boost on AMD hardware, which is why people who overclock already were completely fine with AMD cards when playing the game.

Then there's AC Unity for example. It only had PCSS and HBAO+. Did anyone anywhere actually do any testing with/without these settings to find out if the game's only gameworks effects that ran on AMD were actually dropping AMD performance? It's also an extremely crowded game with most likely a ton of draw calls, did anyone investigate whether AMD's poor DX11 draw call performance was affecting anything?

And the answer to this of course is that no one ever investigates this stuff. Most of the time GW is easy to test, just turn the effect off and see the performance difference. But no one wants to do that for some reason.

Nvidia sponsored titles running worse on AMD at launch does not mean that gameworks is sabotaging or even negatively effecting AMD performance.

In order to determine whether GW is actually harming AMD in a game you have to investigate the GW parts of it.

 

[Good_fella]

Yes it definitely does :whiste:

 

[Gikaseixas]

Nvidia sponsored titles running worse on AMD at launch does not mean that gameworks is sabotaging or even negatively effecting AMD performance.

If this happened on non GW games I would agree with you but it does not as both AMD and Nvidia usually hit the ground running in most games.

I think NVidia gets positive exposures from games reviews as most people are naïve enough to generalize and think that AMD sucks across the board. We as enthusiasts should not think like nubbies, we should know who's the big bad wolf in this story.

 

[RussianSensation]

Yes it definitely does :whiste:

What is this? Marketing slides from AMD/NV do not constitute unbiased data; we need 3rd party reviews or personal data you have recorded. Also, because these marketing slides not only exaggerate the differences between the 2 brands, but because they are depicting data fixed at a certain point in time, they do not account for improved performance in those title via driver updates for either AMD or NV. Pretty much all of those games you listed run perfectly fine on NV hardware today, as well as AMD's. It has already been explained in this thread that certain AMD GE titles do run better on AMD because it does take NV a bit of time to fully optimize the drivers. Conversely, nearly every GWs title "mysteriously" runs worse on AMD, not just at a launch but for months at a time.

Another one is project CARS, it immediately got "lel gameworks doing its thing" reputation but no one will tell you which parts of the game actually feature gameworks, if those will affect AMD performance badly etc. [/B]

If AMD's drivers were so awful and their DX11 performance so atrocious, how come AMD's cards run vendor agnostic and AMD GE titles so well?

I am not seeing anyone providing concrete examples of vendor agnostic games that bomb on AMD's cards. What are these games?

Project CARs is just a messed up game today. For Project CARs, there is absolutely an AMD driver issue since the perfomrance goes up 40-50% moving from DX8.1 to DX10. Still, it's impossible to explain how 780Ti easily keeps up with a 970 in many vendor agnostic titles not under the GW's umbrella, but in Project CARS a 1024 CUDA core 960 is just 6% within a 2880 CUDA core 780Ti.
http://www.computerbase.de/2015-05/project-cars-guide-grafikkarte-prozessor-vergleich/2/

It's also interesting how Kepler performs well vs. Maxwell in games NV doesn't care to market much. GW doesn't just hurt AMD cards but Kepler also performs poorly in many GW titles but runs brilliantly in vendor agnostic titles (780Ti vs. 970). Coincidence?

780Ti is as fast or faster than the 970, how it should be, but that's because NV had nothing to do with any of these titles. Interesting, isn't?

Are we supposed to believe that Maxwell has some magic sauce that in GW titles Maxwell crushes Kepler? I am not buying it. Lots of other new games come out and 780Ti matches or even beats a 970.

 

[Spanners]

Double post my bad.

 

[Spanners]

Yes it definitely does :whiste:

Not a single GW title in those marketing slides! Tu quoque doesn't make anything Nvidia is doing (and the topic of this thread) any better or worse. :whiste::whiste::whiste:

 

[decoy11]

Nvidia's GameWorks isn't going to go away you should all just live with the idea that proprietary middleware is here to stay. Just check out this thread over at UnrealEngine

https://forums.unrealengine.com/showthread.php?53735-NVIDIA-GameWorks-Integration

These aren't AAA developers but indie devs that are eating up every scrap of help they could get from Nvidia and their GameWorks program.

oh and guess what Nvidia is working on

Multi-bounce VXGI multiple indirect bounces (right column) and with single bounce for comparison (left column)

The reactions from these people are all praises and begging for a quick release.

 

[AtenRa]

Nvidia sponsored titles running worse on AMD at launch does not mean that gameworks is sabotaging or even negatively effecting AMD performance.

A few samples,

BF Hardline, GE
GTX780Ti and GTX980 are a little faster than R9 290X. GTX980 is only 12% faster than R9 290X

Evolve, Game Evolve
GTX980 is only ~7% faster than R9 290X

DiRT Rally, GE game
GTX980 is only ~9% faster than R9 290X

Trine 3, neutral
GTX980 is only ~14% faster than R9 290X

Dragon Age Inquisition, Game Evolve with Mantle
GTX980 is only ~12% faster than R9 290X.

Now lets see some of the latest GameWorks games,

Project CARS,
GTX980 is ~22% faster than R9 290X . Also GTX970 is faster than 290X.

Dying Light
GTX980 is 39% faster than R9 290X . Also GTX970 and GTX780Ti are faster than 290X.

Assasins Creed Unity
GTX980 is 35% faster than R9 290X . Also GTX970 and GTX780Ti are faster than 290X.

WatchDogs
GTX780Ti is 25% faster than R9 290X . Also GTX780 is faster than 290X and GTX770 is very close.

If that is not negatively affecting AMD GPUs then what is it?? If the developer is not allowed to optimize the game for both parties then that affects negatively the AMD hardware in the case of ANY GameWork game.

It is another thing to optimize the game for your Hardware and another thing not allowing the competition to optimize the game for their hardware. Even in Mantle games the NVIDIA GPUs perform as they should be, GTX980 is always faster than 290X in every Mantle or GE game. If GE and Mantle games were developed with the same NVIDIA GameWorks "mentality", the R9 290X would be faster than GTX980 in those games. That never happened.

Not only that, latest GW also negatively affect older NVIDIA cards.

 

[Alatar]

Your "neutral" titles are mostly GE titles with devs that have close ties with AMD... Have you considered the possibility that maybe the GE title performance difference between AMD and Nvidia isn't the right one but is an optimistic one for AMD?

Still waiting for the proof of GW actually doing something negative.