Yaffle
Volunteer Moderator
In case anyone cares the new DirectX12 was announced yesterday:
http://blogs.msdn.com/b/directx/archive/2014/03/20/directx-12.aspx
Which boasts about "console level efficiency" a marketing line I can only assume means that it's much better at dealing with software <-> hardware translation.
It seems to be backwards compatible too, so no need for a new card to take advantage of it.
(For anyone who doesn't know the concept behind DirectX is that it has a standard set of inputs at one end and outputs at the other. A software writer doesn't need to know the specs of the graphics card, as long as (s)he gets the standards right on the DX "in" side DX should do the right thing on the "out" side, so driver writers of hardware aim to write communications between the metal and the DX outputs. Sort of. This is very easy to programme for, but is not efficient as the DX layer has to do shenanigans to get from one end to the other. It's the best way we have of dealing with the myriad of configurations a PC can have. Consoles being fixed don't need this, as long as you know what you are doing you can just deal with the hardware). If you fancy it you can code direct to the hardware, but it will be specific to the hardware you code for.
Anyway - in the red corner here's AMD/ATi's spin on it http://www.amd.com/us/press-releases/Pages/amd-demonstrates-2014mar20.aspx
and then the green corner, nVidia says - http://blogs.nvidia.com/blog/2014/03/20/directx-12/
http://blogs.msdn.com/b/directx/archive/2014/03/20/directx-12.aspx
Which boasts about "console level efficiency" a marketing line I can only assume means that it's much better at dealing with software <-> hardware translation.
It seems to be backwards compatible too, so no need for a new card to take advantage of it.
(For anyone who doesn't know the concept behind DirectX is that it has a standard set of inputs at one end and outputs at the other. A software writer doesn't need to know the specs of the graphics card, as long as (s)he gets the standards right on the DX "in" side DX should do the right thing on the "out" side, so driver writers of hardware aim to write communications between the metal and the DX outputs. Sort of. This is very easy to programme for, but is not efficient as the DX layer has to do shenanigans to get from one end to the other. It's the best way we have of dealing with the myriad of configurations a PC can have. Consoles being fixed don't need this, as long as you know what you are doing you can just deal with the hardware). If you fancy it you can code direct to the hardware, but it will be specific to the hardware you code for.
Anyway - in the red corner here's AMD/ATi's spin on it http://www.amd.com/us/press-releases/Pages/amd-demonstrates-2014mar20.aspx
and then the green corner, nVidia says - http://blogs.nvidia.com/blog/2014/03/20/directx-12/
![[VR] Mad Mike](https://forums-cdn.zaonce.net/data/avatars/m/27/27910.jpg?1553527108){kind=avatar}
