AMD Supports Possible Lower Level DirectX
AMD responds to talk about DirectX and OpenGL being lower in the software stack.
AMD is currently responding to reports that DirectX and OpenGL may allow low-level hardware access. An AMD rep says that the company supports and celebrates a direction for game development that is aligned with AMD’s vision.
The response stems from session descriptions pulled from the DirectX and OpenGl presentations that will take place in GDC 2014. According to one session on Direct3D, presenters will discuss future improvements in Direct3D that will allow developers an “unprecedented level of hardware control and reduced CPU rendering overhead across a broad ecosystem of hardware”.
“You asked us to do more,” the DirectX session reads. “You asked us to bring you even closer to the metal and to do so on an unparalleled assortment of hardware. You also asked us for better tools so that you can squeeze every last drop of performance out of your PC, tablet, phone and console.”
For OpenGL, we have this session description: “Graham Sellers (AMD), Tim Foley (Intel), Cass Everitt (NVIDIA) and John McDonald (NVIDIA) will present high-level concepts available in today's OpenGL implementations that radically reduce driver overhead--by up to 10x or more. The techniques presented will apply to all major vendors and are suitable for use across multiple platforms.”
In a way, AMD’s Mantle has pushed the Khronos group (OpenGL) and Microsoft (DirectX) to take the “lower level” route. However the question is: what will happen to Mantle then? Will developers no longer need Mantle with DirectX and OpenGL hovering at the same level in the software stack?
“AMD would like you to know that it supports and celebrates a direction for game development that is aligned with AMD’s vision of lower-level, ‘closer to the metal’ graphics APIs for PC gaming,” reports an AMD rep. “While industry experts expect this to take some time, developers can immediately leverage efficient API design using Mantle, and AMD is very excited to share the future of our own API with developers at this year’s Game Developers Conference.”
“We’ll be sure to share more news and detail with you closer to GDC,” the rep adds.
What will be interesting to see is where all three will go from here. Will Mantle and DirectX somehow merge ideas so that AMD gamers still get the Mantle benefits when using DirectX to run a game? As AMD points out, we’ll find out more at GDC 2014.

Except OpenGL is about 10x slower, because OpenGL is nearly the same speed as DX11. But hey, if you think 10x slower is fine, you can use a lovely 300mhz CPU instead of a 3ghz.
There is more to it than that. There is also the fact that back then when a game crashed it also could potentially and normally took the OS with it. The API route helps keep that from happening, which is why now you will see a "AMD/NVidia driver stopped working and was recovered" message instead of the system BSoDing and rebooting.
Not even Mantle has true direct hardware access as it is still a API which the game calls the hardware through and still masks it. It does offer more of a low level access than DirectX or OGL do but it is nothing like it used to be when games could directly access hardware.
That will never change. Could you imagine the uproar of the people if a game comes out with a bug or glitch that causes constant reboots when launched? Now if a game does that then you know something else is probably wrong such as a bad GPU or bad RAM but if it had direct access figuring that out would be harder.
Here is how I have always and always will see it. Microsoft will retain the majority of PC market share as will DirectX. The main reason is support. You pay for a license of Windows and you get patch support along with technical support.
The biggest problem with things like OpenGL is that they may not update it to support some of the newest features. Microsoft will always be updating DirectX because of the consoles and because they want to keep that grip on gaming as it is very profitable.
As for the article, it is pretty old and honestly the guy is full of it. Both OGL and DX have mostly the same capabilities and both are working on future capabilities. It is up to the company developing the game to decide if they want to jump on DX or OGL. Problem with OGL is driver support is still lacking. To this day I have to actually copy a certain version of the AMD OGL driver into a folder for Doom 3 or RAGE to work properly while with any DX game I just play it.
As for the Mantle/DX merge, it wont. What most likely will happen is DirectX will have their version as will OpenGL and Mantle will disappear or become something else. Microsoft and the guys behind OGL have been around for a long time doing this, I think they wont have an issue.
Square Enix and DICE obviously had some incentive from AMD to do Mantle. Right now Mantle has a very small market share as it only runs on a handful of GPUs (GCN only) so it is not a massive selling point. And with the price inflation for said GCN cards, they wont sell enough.
Either way I think it is great that Mantle did what AMD wanted and pushed MS and OGL to add support for it. I wouldn't mind a slight performance boost. I just hope it gets better for high end CPUs as right now there isn't much incentive to use Mantle over DDX if you have a high end CPU.
With all the good things said about OpenGL, I am mostly using DirectX myself. And the reason for that has nothing to do with performance. DirectX has better documentation, better drivers, better (and available sooner - eg. vertex buffers) features, better debugging, better tools and better support (=> more stable apps with less man-hour investment). Another thing was that in the past, OpenGL was simply bad - like the GLSL (OpenGL shading language) - the stuff of nightmares (it's OK now).
All these points are also applicable to Mantle with one addition - in its current state, it only supports a small fraction of the market (and I do not really expect any changes in that regard in the near future). Furthermore, Mantle does not even compete with Direct3D nor OpenGL - in fact it can't, because it does not (and is not going to, AFAIK) support older HW. It is more like a complementary API - something you implement in addition to OpenGL or Direct3D (a lot of devs simply can't afford to allocate man hours for this).
With all the good things said about OpenGL, I am mostly using DirectX myself. And the reason for that has nothing to do with performance. DirectX has better documentation, better drivers, better (and available sooner - eg. vertex buffers) features, better debugging, better tools and better support (=> more stable apps with less man-hour investment). Another thing was that in the past, OpenGL was simply bad - like the GLSL (OpenGL shading language) - the stuff of nightmares (it's OK now).
All these points are also applicable to Mantle with one addition - in its current state, it only supports a small fraction of the market (and I do not really expect any changes in that regard in the near future). Furthermore, Mantle does not even compete with Direct3D nor OpenGL - in fact it can't, because it does not (and is not going to, AFAIK) support older HW. It is more like a complementary API - something you implement in addition to OpenGL or Direct3D (a lot of devs simply can't afford to allocate man hours for this).
I that direct call causes an unhandled exception what will happen? Will the PC BSOD? Because if not, that's not close enough to the metal. OGL plugins are still nested within OGL, it's kind of like saying that because I can write bits of C code in .NET, that .NET is suddenly capable of programs that are faster than your competitor hat is using C and ASM.
Remember Mantle is not just talking to the driver, it's talking to the driver in the hardware's own language, and no translation has to be done at all.