Oh yeah, I can sense the mobilization of the troops...

I have no idea if this is a good or bad thing. On the face of it, looks positive. But if you consider some of the horrendous gfx glitches and problems some Gameworks titles have had, with Nvidia actively supporting, you have to think it needs a lot of support.
You could get some developers ramping in too many or conflicting effects without the required QC. Gameworks, laughably ironic, benefit is support in development, can AMD offer the same?
Regardless, I can see the draw for devs, given the hardware inside all consoles. Just hope its utilised well and if it is and it hamstrings Nvidia, well, that'll be an interesting 'poo storm'.

the54thvoidI have no idea if this is a good or bad thing. On the face of it, looks positive. But if you consider some of the horrendous gfx glitches and problems some Gameworks titles have had, with Nvidia actively supporting, you have to think it needs a lot of support.
You could get some developers ramping in too many or conflicting effects without the required QC. Gameworks, laughably ironic, benefit is support in development, can AMD offer the same?
Regardless, I can see the draw for devs, given the hardware inside all consoles. Just hope its utilised well and if it is and it hamstrings Nvidia, well, that'll be an interesting 'poo storm'.

At least in most titles, you can turn the effects off. Or at least most of them. Batman and Witcher seemed to have options to kill them entirely. I'll probably do the same for AMD related effects.

the54thvoidI have no idea if this is a good or bad thing. On the face of it, looks positive. But if you consider some of the horrendous gfx glitches and problems some Gameworks titles have had, with Nvidia actively supporting, you have to think it needs a lot of support.

That's probably the crux of the matter. By dumping in the lap of game developers, you have to trust the devs to 1. Actually spend time, effort, and cash to integrate the features, and 2. Get the QA right. Given how many big studio titles are launched in a half-assed state with mega-patches almost the norm, I'm firmly in the " Believe it when I see it" camp. Game developers can't seem to get a polished game out using some pretty ancient game engines that are a known quantity. I don't have much faith that giving them more responsibility - either through features, or API, will improve the situation unfortunately.

the54thvoidRegardless, I can see the draw for devs, given the hardware inside all consoles. Just hope its utilised well and if it is and it hamstrings Nvidia, well, that'll be an interesting 'poo storm'.

Should be enough in the features to satisfy the big studios - whose end credits take longer to roll than the game does to play, but I'm not so sure about the smaller studios - those with more likelihood of producing an actual interesting game that isn't just a re-skin of last years corridor shooter.

AMD will still create these frameworks, but they will be open source, meaning anyone will be able to use them in any way they like. Unlike NVIDIA's GameWorks which is a proprietary framework with limited "functionality". You can use it to easily create a certain effect like hair or water, but you can't really modify it to achieve something a bit different or for a different purpose. Or different hardware...

about time! good work AMD! Although myself would have liked to see this being part of extended Vulkan or DX API for that matter... but good, that AMD finally woke up.

I think its great that AMD is going open source with their alternative to gameworks, Now its all up to the developers to take advantage of it, even without having to pay money.

We still have to remember that pc gaming is the underdog in the gaming world, and since AMD have 100% consoles then it might fly faster than any of us really think which means gameworks will be left behind in the dust, just like Nvidias Physx.

Every time i see some thing like this all i hope is that it shifts market shares closer to 50/50, seeing as that will benefit me most, i am fanboy of myself.

EboI think its great that AMD is going open source with their alternative to gameworks, Now its all up to the developers to take advantage of it, even without having to pay money.

We still have to remember that pc gaming is the underdog in the gaming world, and since AMD have 100% consoles then it might fly faster than any of us really think which means gameworks will be left behind in the dust, just like Nvidias Physx.

Main reason is for features to become core element. For as long as hardware accelerate physics do not become a norm, we won't see game changing physics beyond just eye candy.

CPU physics took a while but since it works the same on all CPU's, games like Half-Life 2 could use it for puzzles or just more realistic feel of the entire game (floating corpses and objects, zombie splitting and fragmentation, ragdolls, objects collision and friction etc). PhysX on the other hand is only accelerated by one type of GPU (NVIDIA). Meaning developers cannot and will not afford losing customers by creating a NVIDIA only game. Especially not today when AMD is holding 100% of the console market. Meaning physics won't evolve anywhere. However, since GPUOpen is open and runs on all GPU's, we might see a breakthrough. We might see new unique games where physics are game's core element. Something you can't create with CPU physics.

RCoonAt least in most titles, you can turn the effects off. Or at least most of them. Batman and Witcher seemed to have options to kill them entirely. I'll probably do the same for AMD related effects.

You can turn off hairworks. Gameworks itself being the SDK is what makes the entire game tick, you can't turn that off. It's like saying you can turn off UE4 on a UE4 engine game.

this is going to be great.

AquinusYou can turn off hairworks. Gameworks itself being the SDK is what makes the entire game tick, you can't turn that off. It's like saying you can turn off UE4 on a UE4 engine game.

Not really. Frameworks are predesigned functions to achieve a specific result. HairWorks is only concerned with realistic hair animation. If you turn it off, entire game will work just fine. You'll just have static hair or less realistic. Which is no the same as removing a part of a game engine.

Though, engine itself can provide such framework by itself and be fully independent. Unreal Engine has bunch of stuff that helps you create content faster (like terrain and props manipulation). Stuff like HairWorks or TressFX is even further more specific feature.

Imagine these frameworks as modules that you add to a core functionality. It's easier to add pre-made modules than build them from ground up from scratch. That's why they exist.

Doubt AMD could have done anything but Open Source. One they have a history of doing it and two, they aren't big enough in market share to be picky.

one word: sweet!
once the bullet 3 gpu pipeline hits a stable state we will finally have a completely opensource alternative to gamesometimesworks.
i just hope developers start to use this, integration into a ready engine like ue4 or pyrogenesis could get the ball moving in the right direction

HumanSmokeThat's probably the crux of the matter. By dumping in the lap of game developers, you have to trust the devs to 1. Actually spend time, effort, and cash to integrate the features, and 2. Get the QA right. Given how many big studio titles are launched in a half-assed state with mega-patches almost the norm, I'm firmly in the " Believe it when I see it" camp. Game developers can't seem to get a polished game out using some pretty ancient game engines that are a known quantity. I don't have much faith that giving them more responsibility - either through features, or API, will improve the situation unfortunately.

It says "developers" but I don't think it means the developers you mean.
There are developers who develop frameworks such as "Unity", "Unreal" etc. (most are multi-platform, including Xbone, PS4 where CGN reigns supreme)

Most of the other developers just use what guys mentioned above have developed. They don't really go too deep into "to the metal" levels.

The best part of this initiative is that it could be supported on nVidia, also barring proprietary "let me leverage my market position" things like Gameworks.

Damn, GPUOpen not released yet but 53% already hyping it.

RejZoRAMD will still create these frameworks, but they will be open source, meaning anyone will be able to use them in any way they like. Unlike NVIDIA's GameWorks which is a proprietary framework with limited "functionality". You can use it to easily create a certain effect like hair or water, but you can't really modify it to achieve something a bit different or for a different purpose. Or different hardware...

This opensource myth must die. How many games using TressFX, Mantle, and other stuff? 5000-6000?

RejZoRAMD will still create these frameworks, but they will be open source, meaning anyone will be able to use them in any way they like. Unlike NVIDIA's GameWorks which is a proprietary framework with limited "functionality". You can use it to easily create a certain effect like hair or water, but you can't really modify it to achieve something a bit different or for a different purpose. Or different hardware...

trying to debug code that uses blackboxes is a pita... if you can drop behind the curtain and see what is happening you can fix your shit.

There has been a lot of debate on this on many sites over the last couple days, and the general consensus is: Good going AMD, finally making good on things that should have been delivered years ago to the devs, community, and to show a unifying support for everyone, and hell.... it could have prevented some of the issues we had and have.


That being said, it needs a major kick in the pants to get moving, so they need to throw some hardware at people to ensure that some community devs will use it and get some awesome payback for it. Make X demo or game on any GPU and we will give you a new free GPU for the next two generations as long as your project works and source code is made available for others to build on.

RecusDamn, GPUOpen not released yet but 53% already hyping it.

Spoiler

Hypes so far:

Fusion failed.
Bulldozer failed.
Mantle failed. (HD 7850 w/ Mantle will be faster than Titan)
Fiji failed. (40% faster than Titan X)

This opensource myth must die. How many games using TressFX, Mantle, and other stuff? 5000-6000?

Man, look at the bright side of things:

FUSION= APUs, since AMD put they on the market, Intel increased their R&D on GPUs thorough the roof.
Bulldozer= Focus on multi thread computing, maybe not the best IPC, but does what promises for really affordable price point.
Mantle= Low level access for GPUs, without Mantle theres is no DX12 or Vulkan.
Fiji= Introduction to HBM, Fury X was meant to be on top of the GTX 980, if Nvidia released the 980ti 2 weeks after the FuryX, the asking price would be lower ($550~600, just like the R9 290X).
opensource myth must die??? Are crazy? Do you understand what are you saying? No comment here.

Look man, tell me, why you buy Nvidia cards?

*The best value card on the high end market right now is the 390X, nV offers are overpriced.
*AMD cards output higher color bits, Radeon HDRI/HDR engine is by Native 128bit, but nV is 24bit.
*D3D11 rendering on AMD cards is true 32bit. since the GTX700 nV only uses 24bit.
*AMD have all around better optimized AA, nV FXAA is a joke.
*FreeSync Monitors are more cheap and deliver a better result, and you have choice, Gsync begs for you to turn on Vsync in your games, instead AMD give you a frame limiter that don`t mess with your input lag.
*nV don`t care about VR, AMD develop and support LiquidVR.
*AMD don`t release 3.5Gb +500Mb cards and advertise like it has a full 4GB (GTX970 I am looking at you).
*AMD don`t release drivers to lower performance on older cards just to promote the new line up.
*And they sure don`t release and promote proprietary code that crash games (GameWorks)
*But nVidia does great things to!
So put some faith on the AMD guys, they are apparently the only ones pushing PC gaming forward or they have a really good PR department.

This image never gets old because NVIDIA keeps proving it true:


I applaud AMD, again, for doing the right thing.

the54thvoidI have no idea if this is a good or bad thing. On the face of it, looks positive. But if you consider some of the horrendous gfx glitches and problems some Gameworks titles have had, with Nvidia actively supporting, you have to think it needs a lot of support.
You could get some developers ramping in too many or conflicting effects without the required QC. Gameworks, laughably ironic, benefit is support in development, can AMD offer the same?
Regardless, I can see the draw for devs, given the hardware inside all consoles. Just hope its utilised well and if it is and it hamstrings Nvidia, well, that'll be an interesting 'poo storm'.

I think it's a good thing. However, Like @RCoon said, as long as you can turn them off. Hell, IMHO, GameWorks hasn't worked out real well for Nvidia either, with most Nvidia owners also having to turn either most or all the package off in order to max out textures and such.

@Aquinus, actually, in Arkham Knight, for instance, you can turn every one of the elements off or on individually for the Gameworks package. The settings even point out which are Gameworks. Turning them all off except for the special light rays is what allowed me to crank up the settings and play the game smoothly without a crash.

so glad i jumped to amd, got sick of putting up with nvidia proprietary closed nature. go AMD for open source!!!

This is a good news. But let's hope it's not just another AMD-Pixelux's Open Physic Initiative.

I'd be excited if they'd made their drivers and all related software open source.

okidnaThis is a good news. But let's hope it's not just another AMD-Pixelux's Open Physic Initiative.

This has been exactly the kind of vaporware that AMD/ATI have been guilty of releasing, and you are right they need to build initiative for this new attempt by putting their money out there, or by just building more than the most rudimentary tools and then moving on to other pet projects.


Mantle was a significant change from their past attempts, as was TressFX, and frame control tech. I hope they continue, for their own sake, many of us have long memories and are tired of getting crapped on with their failed attempts.