I would say AMD and Nvidia better support OpenGL 4.2 as best as they can instead. Of course writing directly to hardware is cool, just because they are only two major video graphics companies left standing. So this announcement gets in "mixed bag" category for me. However it just shows how much Microsoft really cares about PC gaming.
You are right, this is nowhere near as bad as the crudfest that Glide was.
We could go around in circles on this, but really we are just looking at an AMD specific layer in the cake, you can make of it what you will.
Personally I hope this will cause a paradigm shift in the whole industry and get both Microsoft and nVidia to follow suit and offer something similar.
We could go around in circles on this, but really we are just looking at an AMD specific layer in the cake, you can make of it what you will.
Personally I hope this will cause a paradigm shift in the whole industry and get both Microsoft and nVidia to follow suit and offer something similar.
Well according an acquantaince of mine that unlike me actually knows something on the matter, "Open GL is simply "substandard" for game development. Putting lipstick on a pig comes to mind. DX is better, but windows only"
While he may be biased, the fact is that OGL was not born for games.
...but it is great for CAD applications and line drawings
Indeed: high quality, very slow moving stuff
First OGL was not born for games, true. Not so true about 3th and 4th versions of OGL. DirectX is better, but it also mean that OGL isn't viable solution outside Windows platform.
This is true. Again it is simply a question of preference.
I.e. What hardware are you targeting? What OS? Mobile/desktop/console? Etc...
Any good developer will tell you that you should pick the tool that is right for the job.
I think it is a matter of names and and being similar, not equal: both consoles have a low level language. Although they are bound to be different, running in the same HW graphic wise (main differences should be in the memory subsystem), they should be pretty similar and follow the same optimization rules.
Mantle, although for PC only, should be similar to both language (maybe closer to the PS4 due to the special memory of Xbone).
At the end what matters is how easily can you convert from one to another and at what performance level.
I'd guess at the following
1. Elite will not support Mantle (it's to late in the process for it. Maybe release 2)
2. Mantle sort of fails and never achieves a great success (because it cuts out both Nvidia and Intel video cards)
3. The basic ideas and performance improvments are built into DirectX 12 and OpenGL (something)
It's a nice idea, but a private low level interface is going to be hard to get support for, unless they open it up to Nvidia and intel as well.
1. Elite will not support Mantle (it's to late in the process for it. Maybe release 2)
2. Mantle sort of fails and never achieves a great success (because it cuts out both Nvidia and Intel video cards)
3. The basic ideas and performance improvments are built into DirectX 12 and OpenGL (something)
It's a nice idea, but a private low level interface is going to be hard to get support for, unless they open it up to Nvidia and intel as well.
Agreed.
Intresting aside here: http://msdn.microsoft.com/en-us/library/windows/desktop/dn312084(v=vs.85).aspx
The tiled resources and independent swap chains are very interesting.
Some related points raised in this video that some may find interesting:
http://www.youtube.com/watch?v=MH2hjhcfWic#t=1h23m13s
http://www.youtube.com/watch?v=MH2hjhcfWic#t=1h23m13s
Being a dev myself I would have to agree with the guys on the panel, after all they are arguablly the top three engine developers in the world.
They say that it would be a bad idea to have yet another API. Instead Microsoft and the Khronos group should take onboard what Mantle is doing to improve DX and GL.
This way we all benefit.
I believe there already is a low level language for nVidia's Geforces, but it is basically ignored (maybe except for some nVidia sponsored games and in a very light maner).
Mantle is quite significant not in technical terms per se, but because it leverages the console ecosystem synergies with the PC. That and the effect of rocking the PC DX stagnation. All good things for PC users.
The problem is although Mantle can impact on DX in order shake MS out of its sleepy state regarding gaming, there will always be the trade-off of being more general but less performant versus being more restrictive and efficient.
In the days of Glide, where architectures were plenty, general won by a mile. But today, the PC space is divided in three different players (although only 2 are relevant for gaming). It is in the mobile space where the fragmentation exists.
Another issue is something I saw referred, but can't really coment on (for lack of knowledge): DX (and Open GL in a worse way) are based upon ideas and fundamentals brought from an old era and have become completely out of sync with todays game engine design rules and way of functioning. That a fundamental rewrite of the standard is needed.
