Mehca :
Lots of people keep saying that Vulcan is some kind of future API that will be widely adopted and this AMD is king and Toms is guilty of Nvidia conspiracy. Oddly enough I remember this exact thing being said about Mantle, yet Mantle didn't see widespread adoption, and didn't improve performance nearly as much as people claimed. Vulkan is basically Mantle 2, and I really don't see it getting widespread adoption either, even with AMD having the market on console hardware.
DX12 is the future, and you can rest assured Nvidia will perform quite on that API, once it actually becomes relevant. The vast majority of games out there are DX11 though, and most games that support DX12 will also support DX11. DX11 is still the most relevant API, and is where Nvidia is rightly investing its support. DX12 optimization will be given through drivers, which Nvidia has a pretty good track record for.
Fanboyism won't change reality.
Reality is DX12 is heavily based on Mantle- just as Vulkan is.
The main advantage of Vulkan over Mantle is that it does not have the hardware limitations of Mantle- it is not tied to any specific hardware (this is also true for DX12 of course). Where Vulkan has a huge benefit over DX12 however is that it isn't tied to a single software platform. DX12 is Windows 10 ONLY. Vulkan is the successor to OpenGL and is available on Windows 7, 8, 8.1, 10, Linux (all modern distros), MacOS, iOS and android (there are probably more I can't think of right now).
The reason DX always had much higher market share in the past was it was easier to develop for and better supported / documented than OpenGL, and there wasn't really much pressure for cross platform support.
This gen I think Vulkan could be much more relevant for a few reasons:
1: The largest Windows install base is Windows 7, developers cannot address that market if they go DX12 (so currently most DX12 games also include a DX11 render path- however I think longer term a Vulkan based renderer would address all Windows users with one renderer, rather than developers having to support 2).
2: Vulkan and DX12 are much more similar than DX11 and OpenGL- as they share a common base platform so moving over to Vulkan shouldn't pose the same difficulties as going from DX -> OpenGl did.
3: The wide compatibility with other software platforms as noted above
4: The fact that there are some serious hitters in the industry who are pushing for games to come to Linux as a platform (specifically Valve and Steam OS). There have already been a number of high profile OpenGL ports done to bring big titles to SteamOS, Vulkan should make that process much easier longer term, and won't result in giving up on performance if the new DOOM is any indication (although it is worth noting that ID have always developed for OpenGL and are probably the best in the industry at it so DOOM is probably the very best performance we can expect).
The lack of adoption is mainly down to the fact that Vulkan was only just released whilst DX12 has had some time on the market. It will be interesting to see where things go.
I think the key thing to remember though is *even if* developers move over to Vulkan instead of DX12, the two APIs offer essentially the same improvements in terms of performance and better hardware utilization. The thing is though if you look at the overall picture one of two things happens with either- games either perform effectively the same under DX11 and DX12 *or* AMD gain more from the switch than nVidia do. I've yet to see a title where nVidia gains more FPS than AMD, and I think that is simply down to the fact AMD have dedicated hardware built into their gpu's that nVidia don't. That said- as nVidia are generally ahead in DX11 performance (for a given performance bracket of card), what this is doing for the most part is *leveling the playing field* between the two. The reason this is exciting on AMD cards is because of the weakness in DX11 the cards are priced lower than they probably should be- so they currently represent a good deal. I mean nvidia cards can run DX12 and Vulkan titles just fine, they aren't losing performance.
What I do think is that in light of this, nVidia will probably have to adjust prices down (and AMD will likely nudge prices up) as DX12 / Vulkan become more common. I don't believe nVidia can get the kind of gains AMD can purely though drivers as they are fighting hardware with software tweaks (nVidia are adding in hardware async with Volta in 2018). I mean based on theoretical performance the R9 Fury is a much more powerful card than a GTX 980, yet in DX11 they are roughly on par. In DX12 / Vulkan we are seeing the Fury positioned above the 980- where it frankly should be, whilst more powerful nVidia cards like the 1070 are sitting above it.