Vulcan Delayed

[3DVagabond]

So there's no advantage to being able to run Win 10, just like DX12 does and also being able to run additional OS's that DX12 can't? Seems like pretty simple reasoning to see how you are wrong.

 

[dark zero]

Irrelevant. Windows 10 is a free upgrade for 7+. Anyone running an older version shouldn't be doing so as it no longer has (mainstream) support.

Seriously, Vista is ten years old. Somebody that has hardware that can handle future Vulcan games won't be running a ten year old OS.

And again, we have 2 Wolfenstein games now and Doom as upcoming. I'm not sure why anyone is expecting Vulcan to suddenly create a landslide of Windows games.

I'll bet there'll be less Vulcan games than hardware PhysX games, and the latter is already a laughable niche.

No is not. Windows 10 was never a free upgrade. It only last a year and also you are forced to download every update even if it is trash.

Also Vulkan helps Windows 7 and 8 users thanks to Windows 10 bad way to manage the updates (Windows 10 was great at first, but after those 2 bad updates, I had it, but then returned to Windows 8.1 for a while due the ustability of the OS). It helps Linux and even MAC users. Also Steam and Origin gets benefited with that and unless MS buys them, they can do whatever they want with their products. Also seems that their Steam OS project is still alive. And they knows that MS is on a similar project now.

 

[96Firebird]

Windows 10 was never a free upgrade. It only last a year and also you are forced to download every update even if it is trash.

This is wrong. The offer is only available for a year.

 

[Spjut]

Windows 10 being a free upgrade for one year obviously gives DX12 a better chance than DX10/11 had, but if the companies don't think the adoption rates are high enough, it still ultimately won't matter

Vulkan undeniably has the advantage in that once it's out, it can target the userbases of Win7 and later from day one, instead of waiting for enough to move to Win10

 

[ShintaiDK]

It was the same story with DX10 and DX11 (And before). OpenGL started to support the same features. But we all know how it went. It wont be any different with Vulcan.

DX wins every time. Also the Xbox is DX. While the PS4 is GNM/GNMX. So no help for Vulcan there.

 

[TheELF]

Windows 10 being a free upgrade for one year obviously gives DX12 a better chance than DX10/11 had, but if the companies don't think the adoption rates are high enough, it still ultimately won't matter

This depends,since every game is a console port and in theory console games already run with the mechanics that Dx12 will use implementing Dx12 in every game might be as easy as clicking a checkbox come compiling time,if so every game will support Dx12 and if vulcan is just as easy to implement than every game will have that too.
It all comes down to money for the devs,if they can put it in without much effort (workhours=money) they will put it in if only to be able to boast about having it.

 

[dogen1]

It was the same story with DX10 and DX11 (And before). OpenGL started to support the same features. But we all know how it went. It wont be any different with Vulcan.

DX wins every time. Also the Xbox is DX. While the PS4 is GNM/GNMX. So no help for Vulcan there.

Hasn't it already been said that it's not the same situation as with openGL? Many of the reasons that devs don't use OpenGL don't apply to vulkan at all. They're not comparable.


Anyone who thinks less games will use Vulkan than GPU Physx, or even OpenGL on PC must have forgotten that Unreal 4, Unity, CryEngine(most likely) and Source 2 are all guaranteed to support Vulkan, meaning that any linux versions(and possibly the windows versions) of games using those engines will use it. In the case of source, all valve games will use it going forward, possibly to the point of replacing D3D.

 

[zlatan]

I think Vulkan is a really good option for the future. SPIR-V is much better than D3D bytecode, and Khronos did a very good job. D3D12 is super, but the API is really designed for Xbox One, and some points it heavily favors GCN. For example the multi engine concurrency is extremely bad with Maxwell and kepler. Vulkan has some useful options to improve the performance on Maxwell and Kepler, so this API is a real multi-vendor API, and not just gives the ability to NV and Intel to support the important features efficiently in their future hardware designs.

Vulkan is also more simple, because it just have command buffers, so the API is using a flat procedural approach, which is much better for Intel integrated GPUs.

Of course D3D12 has some unique features, but Vulkan is more efficient and more vendor natural. I think this is more important than some fancy features.

 

[3DVagabond]

This depends,since every game is a console port and in theory console games already run with the mechanics that Dx12 will use implementing Dx12 in every game might be as easy as clicking a checkbox come compiling time,if so every game will support Dx12 and if vulcan is just as easy to implement than every game will have that too.
It all comes down to money for the devs,if they can put it in without much effort (workhours=money) they will put it in if only to be able to boast about having it.

I have no doubt that the port from Xbox1 will be simple. Sony uses a different API though. What they have in common is GCN. To that point though, We were told by the devs that adopted Mantle that adding an API isn't as big of a deal as some people think. They already write their games for several API's anyway.

 

[Spjut]

IMO, the DX12 and Xbox One argument sound logical, but at the same time, shouldn't the Xbox One already have been driving DX11.2 adoption as well?

AFAIK, Frostbite 3 is the only engine to support DX11.1, and have shown increased CPU performance in benchmarks, whereas all others are still stuck at DX11.0

 

[TheELF]

Personally I don't expect to see many dx12 specific features for a very long time,games will be Dx12 for the close to the metal part (no driver thread) and probably for the texture streaming part and if we are lucky for the draw call part as well and other than that business as usual they will be Dx11 games for all the effects and anything else.
Devs spend too much time and money on optimizing for Dx10/11 to just move ahead.

 

[3DVagabond]

Personally I don't expect to see many dx12 specific features for a very long time,games will be Dx12 for the close to the metal part (no driver thread) and probably for the texture streaming part and if we are lucky for the draw call part as well and other than that business as usual they will be Dx11 games for all the effects and anything else.
Devs spend too much time and money on optimizing for Dx10/11 to just move ahead.

They spend a lot of time optimizing for the consoles too. That should translate over. I'm personally shocked that there still aren't any DX12 games. Has this ever happened before where we've had a new higher performing API and months after release still nobody's using it?

 

[TheELF]

They spend a lot of time optimizing for the consoles too. That should translate over.

That's what they tell us, but they use PC game engines in the games and I'm pretty sure they do the game design on PC as well ,which means they use Dx11 and somehow translate it over to the APIs.


A lot of games still are Dx10 and even Dx9...

 

[ShintaiDK]

They spend a lot of time optimizing for the consoles too. That should translate over. I'm personally shocked that there still aren't any DX12 games. Has this ever happened before where we've had a new higher performing API and months after release still nobody's using it?

Yes, took quite a while after DX10 for example.

 

[ShintaiDK]

That's what they tell us, but they use PC game engines in the games and I'm pretty sure they do the game design on PC as well ,which means they use Dx11 and somehow translate it over to the APIs.


A lot of games still are Dx10 and even Dx9...

Not to mention a lot of console games are higher API based. GNM and GNMX both exist for a reason for PS4 for example. One could call them DX12 and DX11.