• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Nvidia’s GameWorks program usurps power from developers, end-users, and AMD

So turn off HBAO. Or are you suggesting nVidia should spend time and money on a project for AMD?

I'm suggesting Nvidia allow AMD to optimise. This is how its always worked...until now.

bzzzt, not according to the article

The article is wrong. AMD cannot optimise for GW HBAO. They can still use it they just can't optimise for it, unlike Nvidia.

If im wrong on this Joel please correct me.
 
Last edited:
How about the game developers just stand up to the Green Goblin for once and demand open standards?

Its a difficult one to draw the line, in many respects using GW's ambient occlusion system, etc. isn't much different from what would happen if a company decided to use for instance enlighten if a company doesn't have the resources or budget to implement a feature like that from scratch but they can get it from another source free or cheaper than doing it all for themselves then they are likely to do so and unless the full source is licensed neither the developer or GPU vendors can see inside the library code easily.

Though as Andy pointed out nVidia and AMD can profile their drivers in debug mode and work out enough to do 99% of optimisations it just requires a bit more effort on their part.
 
bzzzt, not according to the article



if HBAO+ doesn't run on AMD hardware then how can it affect AMD performance? simple answer is that it can't

the article says that gameworks means that devs and AMD CANNOT improve performance, yet clearly between the two of them they did


I also don't believe for a single second that AMD only have ASSEMBLER tools for their own drivers, that would be retarded. If you write the drivers in C++ then you can log every single call that is made IN TO your own software from outside - if you did C++ at college then you should know this. Even if the tools didn't already exist within C++ (which they do), you would only have to add a line of code to every function to tell it to log every time it was called and what arguments were passed to it and any other info you wanted to log.

If you log that a particular exe or lib calls in to your exe and asks it to perform X then Y then Z and you see that this is causing a performance issue you can capture that and tell your own application to perform A > B > C instead.

Sorry, can't go into more detail on this situation. I did not say AMD only had assembler tools for driver optimization. I said AMD could still perform assembler-level optimizations based on what they can see from the driver's output, but cannot see the HLSL code and therefore can't optimize at that high level.

In response to "NVidia are playing smart with this at launch"... what happened with Tomb Raider again? oh yeah, the developer / AMD made a load of changes to TressFX in TR *right* before launch meaning that all the launch reviews showed a GTX680 matching a 7870 and Nvidia then had to rush out a few patches to get back that performance

Yes. And NV was able to do that because NV wasn't barred from seeing the source code. TressFX is *available.* You can go download it.
 
The difference between Gaming Evolved / TWIMTBP and GameWorks is that in the former programs (including TressFX), games that launched with preferential support for one program did not preclude eventual optimization for the other vendor.

GW does preclude this by using NV libraries under license that AMD cannot see and cannot optimize in the same fashion. It may be that a developer that wants AMD optimizations can work with the company to get some of that work in. AMD may be able to do some fine-tuning via ultra low-level driver tuning. But the developer cannot hand AMD the HLSL code and say: "Here, have a look." In the past, they could. That's the difference. And because any developer cooperation with AMD to try to get some kind of code optimization in is going to operate under a legal gray area, that's going to be difficult and time consuming.

Each library is specific, not broad. So a library for HBAO+ is a library for that function. A library for tessellation (if one existed) is a library for tessellation. A game that incorporates five GW libraries for five separate functions does not prevent AMD from optimizing for a sixth function that is not covered by a GW library.

It's a simple distinction.
 
Last edited:
They should all work but the will only have been QA'd against nVidia cards. Not sure at what level Flex ties into PhysX so might be some potential issues there but IIRC it has its own specialised solver thats stand alone from PhysX.

Not sure if its been released bur flex will have a full directcompute version in physx 3.4 sdk
 
They should all work but the will only have been QA'd against nVidia cards. Not sure at what level Flex ties into PhysX so might be some potential issues there but IIRC it has its own specialised solver thats stand alone from PhysX.

Not sure if its been released bur flex will have a full directcompute version in physx 3.4 sdk

So if this is the case, the article is because AMD can't optimize for this nVidia tech? This is getting more and more like Mantle the more this goes on but Mantle is cool and GameWorks is bad?
 
The difference between Gaming Evolved / TWIMTBP and GameWorks is that in the former programs (including TressFX), games that launched with preferential support for one program did not preclude eventual optimization for the other vendor.

GW does preclude this by using NV libraries under license that AMD cannot see and cannot optimize in the same fashion. It may be that a developer that wants AMD optimizations can work with the company to get some of that work in. AMD may be able to do some fine-tuning via ultra low-level driver tuning. But the developer cannot hand AMD the HLSL code and say: "Here, have a look." In the past, they could. That's the difference. And because any developer cooperation with AMD to try to get some kind of code optimization in is going to operate under a legal gray area, that's going to be difficult and time consuming.

Each library is specific, not broad. So a library for HBAO+ is a library for that function. A library for tessellation (if one existed) is a library for tessellation. A game that incorporates five GW libraries for five separate functions does not prevent AMD from optimizing for a sixth function that is not covered by a GW library.

It's a simple distinction.

Well put.
 
It is nonsense to say AMD can only do assembler level optimisation. The drivers will be written in C or more likely C++.

If you can profile your own c++ code you can intercept and redirect the c++ calls, it really is that simple.

It is marginally more time consuming to have to do it that way, but it is absolutely not impossible.
 
Last edited:
So if this is the case, the article is because AMD can't optimize for this nVidia tech? This is getting more and more like Mantle the more this goes on but Mantle is cool and GameWorks is bad?

No its not like Mantel because AMD cards still going through some of GameWorks= nVidia tech while with Mantel=AMD tech, nVidia cards dont go through any of it so are not hampered by it or restricted by it at any level.

Its like saying that AMD cards are hampered by CUDA which is wrong because AMD cards dont run CUDA so cant be hammered by it even if NV cards get twice the performance running CUDA, that's still not hampering AMD cards.
 
Last edited:
No its not like Mantel because AMD cards is still going through some of GameWorks= nVidia tech while with Mantel=AMD tech, nVidia cards dont go through any of it so are not hampered by it or restricted by it at any level.

It doesn't matter how many times you say it or explain it. They choose not to listen. You're Just wasting your time. If they can't understand the very basic difference by now, you'll never get them to understand it. It comes down to their gpu preference and they won't hear anything said against it.
 
No its not like Mantel because AMD cards is still going through some of GameWorks= nVidia tech while with Mantel=AMD tech, nVidia cards dont go through any of it so are not hampered by it or restricted by it at any level.

But at least GameWorks is working on AMD cards, whereas Mantle isn't working on nVidia cards. I am not even sure what is what, as not one AMD owner (except Martini) has said how Batman AO runs (and Martini said it was great on his 290).

As to tessellation, AMD should have improved from the 6 series but sadly they didn't, so should we cut out tessellation because MAD cards can't cope with it?
 
It doesn't matter how many times you say it or explain it. They choose not to listen. You're Just wasting your time. If they can't understand the very basic difference by now, you'll never get them to understand it. It comes down to their gpu preference and they won't hear anything said against it.

Pot, Kettle, black?
 
It doesn't matter how many times you say it or explain it. They choose not to listen. You're Just wasting your time. If they can't understand the very basic difference by now, you'll never get them to understand it. It comes down to their gpu preference and they won't hear anything said against it.

Indeed.
 
It doesn't matter how many times you say it or explain it. They choose not to listen. You're Just wasting your time. If they can't understand the very basic difference by now, you'll never get them to understand it. It comes down to their gpu preference and they won't hear anything said against it.

Had to laugh :D :p
 
The difference between Gaming Evolved / TWIMTBP and GameWorks is that in the former programs (including TressFX), games that launched with preferential support for one program did not preclude eventual optimization for the other vendor.

GW does preclude this by using NV libraries under license that AMD cannot see and cannot optimize in the same fashion. It may be that a developer that wants AMD optimizations can work with the company to get some of that work in. AMD may be able to do some fine-tuning via ultra low-level driver tuning. But the developer cannot hand AMD the HLSL code and say: "Here, have a look." In the past, they could. That's the difference. And because any developer cooperation with AMD to try to get some kind of code optimization in is going to operate under a legal gray area, that's going to be difficult and time consuming.

Each library is specific, not broad. So a library for HBAO+ is a library for that function. A library for tessellation (if one existed) is a library for tessellation. A game that incorporates five GW libraries for five separate functions does not prevent AMD from optimizing for a sixth function that is not covered by a GW library.

It's a simple distinction.

Personally I think there are 2 issues with this subject that need to be seperated more (atleast until more information comes to light) a lot of things thread is a bit of a mess of confusing the 2 aspects.

On the one hand GameWorks itself has the whole potential issue of it potentially being used by nVidia to subversively undermine the competition which doesn't benefit anyone - the incidental nature of the libraries I think is a bit of a red herring as while it potentially allows nVidia to abuse them to marginalise AMD its not in itself any different from other libraries where neither the developer or the GPU vendors have access to the full source which isn't an unusual case at all with game development. That nVidia can in this case influence the library code is really a footnote to the bigger potential issue.

Then you have Batman AO - where I believe your article showed much of the performance bias stems from the tessellation implementation and the real interesting bit there is why AMD didn't get to include their optimsations - as you said tessellation as used in the game shouldn't be tied in any way to GameWorks so that raises concerns - especially if it comes to light GameWorks was used as a tool to shutout AMD's efforts to optimise - but at this point thats conjecture.
 
Had to laugh :D :p

I hear and see plenty said against AMD on these forums and have no problem with it. Doesn't always work the same way though. At this stage Nvidia could release a press conference saying the libraries are locked and they will not allow AMD to optimize its own performance/drivers. Yet still you'd have a few people in this thread saying but what about Mantle? :D
 
Back
Top Bottom