***The Official Vulkan API Thread***

[KillBoY_UK]

Its great Vulkan is out there and DX12 will have a 2 heavy hitters in march/april. However lots and lots of readers seem to think a to the metal API will be the magic bullet, I've always argued not everyone will want to use such a API epspecaily in the PC space with so many graphics cards out there and coming out that you would have to program and test for. Croteam put it better them selfs in there Vulkan FAQ.

Q: That's better. Now, why was Vulkan support added to Talos?

A: Good question! We (Croteam) firmly believe that Vulkan is really the best low(est)-level API there can be. Fast and portable!
Yes, it has downside(s). For one, it's quite hard to program for. You have to do a lot of things manually, instead of relying on drivers to do the work for you. This is both good and bad at the same time. Good for performance reasons, because driver doesn't assume what game wants to render (I won't go into any more details here, sorry). Bad because there's a lot more coding and in general, it's a more complex approach. You better know what you're doing, because you won't get any help from the driver. You're on your own.
It's really great to have that much control. IF you know what you're doing!

That being said I hope the bigger and better PC dev's will jump on Vulkan/DX12 as they have resources to do so, the CDred's the Bioware's, the Dice's etc. Smaller teams I would think will end up sticking to DX11 like api.

That's my 2p anyway

 

[CAT-THE-FIFTH]

No one is saying AMD did not greatly contribute. What we are saying is it’s unfair in saying AMD did the majority of the work and that the majority of Vulkan is Mantle. What about the major contributions from the other GPU players?

Yes its true vulkan would not have been released in its current form without mantel as a basis. But the same goes for the major contributions the other GPU players did. Take any one of the major players out like ARM or IMG and Vulkan would have been very different.

Even Anandtech said Vulkan is based on Mantle with relevant parts changed or improved as required:

http://www.anandtech.com/show/10035/vulkan-10-released

We’ll start things off with the Vulkan API itself. Since Khronos’s last major press update almost a year ago in March of 2015, not a great deal has changed on the technical side from a high level. After being gifted Mantle 1.0 from AMD – an action that significantly sped up the development process and bypassed the need to figure out some fundamental questions about how the API should be designed – the consortium went about adapting Mantle to serve as a wider, more generic API suitable for hardware from multiple vendors across multiple OSes.

The end result is that Vulkan has its roots firmly in Mantle, through Khronos has worked to make it very clear that multiple vendors are responsible for contributing IP that ultimately went into Vulkan. And while the specific low-level details of the API are beyond the scope of this article, I do know that the shader resource binding system is significantly different from Mantle, and that’s not the only system that was updated or overhauled during Vulkan’s development.

Basically the basic framework of how Vulkan works is based on Mantle,and has improvements based on input from other vendors. After all it has to work on more than one graphics uarch,so its blooming obvious FFS they would need to change parts!!

Its like with VW group cars - you have Seat,Skoda and VW branded cars sharing the same basic floorplans and engines,but with different bodies,interiors,etc tailored to different markets.

Like the Boeing F16 and the Mitsubishi F-2. The latter is based on the F16 but has big changes,ie, different wings,avionics,etc. But still very much is based on an F16.

In fact in aviation and military stuff,its VERY common to have these technology sharing agreements.

It might be a shock for many hardware enthusiasts on forums,but out there in the real world,there are plenty of case of companies making products based on derivatives of other products and marketed or sold as something different.

 

[Freddie1980]

** Mod edit - Comment removed **

------------------------------

It will nice if Vulkan can gaining traction amongst developers but I really can't see it taking off beyond a handful of token titles and some steam games unless DX12 turns out to be a bit of a turnip. Companies aren't going to spend valuable time and resources coding for two different API's unless it has a compelling reason to use it.

IMO Vulkan best bet for success is to be adopted by indie developers, if they can demonstrate how good Vaulkan can be and it has genuine benefits over DX12 then that might be enough for it to gain mainstream adoption with the large publishing houses.

 

[ICDP]

****

It will nice if Vulkan can gaining traction amongst developers but I really can't see it taking off beyond a handful of token titles and some steam games unless DX12 turns out to be a bit of a turnip. Companies aren't going to spend valuable time and resources coding for two different API's unless it has a compelling reason to use it.

IMO Vulkan best bet for success is to be adopted by indie developers, if they can demonstrate how good Vaulkan can be and it has genuine benefits over DX12 then that might be enough for it to gain mainstream adoption with the large publishing houses.

This only applies to Windows based games though. Vulkan has suddenly given developers on Linux and other (notable exception of Apple) OS systems with a much more viable and easier to work with API. In fact a developer can use Vulkan and ignore DX12 all together and still have a fully Windows compatible game.

Vulkan will not fail based on how good or bad DX12 is IMHO, the fact that both Nvidia and AMD are on board along with Android and Linux puts Vulkan in a very good position. IMHO an even stronger position than DX12 and it's reliance upon Windows.

 

[GoogalyMoogaly]

This only applies to Windows based games though. Vulkan has suddenly given developers on Linux and other (notable exception of Apple) OS systems with a much more viable and easier to work with API. In fact a developer can use Vulkan and ignore DX12 all together and still have a fully Windows compatible game.

Vulkan will not fail based on how good or bad DX12 is IMHO, the fact that both Nvidia and AMD are on board along with Android and Linux puts Vulkan in a very good position. IMHO an even stronger position than DX12 and it's reliance upon Windows.

Couldn't most of that have been said about OpenGL in the past too?

 

[ICDP]

Couldn't most of that have been said about OpenGL in the past too?

OpenGL was marginalised because it is a much harder API for developers to work with than DX. It takes more resources to get the same effect and performance you can get with DX. Time = money

Vulkan removes this hurdle by being easier to work with and being low level allowing easier direct access to the hardware. While it can be argued that DX12 is a lower level API as well it must be remembered it is Windows 10 only. Vulkan opens up a whole new potential customer base for those with Windows XP, Vista, Windows 7, Windows 8, Linux, Android, SteamOS. Conversely it also opens up the potential for us as consumers to ditch the Windows 10 requirement for our future PC gaming.

 

[Gregster]

Historically NVidia have been on the ball with OGL and Linux whereas AMD have been shocking. With Vulkan, AMD are now supporting that, so at least Linux has a chance. HL3 exclusive to SteamOS or Linux would have me dual booting and I am sure many others as well.

 

[Orangey]

Vulkan easier to work with?

It seems the hype train is running away on this one.

 

[Mauller]

Couldn't most of that have been said about OpenGL in the past too?

The problem with OpenGL is that it lacks conformence tests and too many defunct ways to do the same thing. This has left a problem where different methods work better, worse, or not at all on different developers drivers.

That is the main reason for a lack of use of OpenGL.

 

[Mauller]

Vulkan easier to work with?

It seems the hype train is running away on this one.

It works better because there are conformance tests and well defined ways of using vulkan compared to OpenGL.

OpenGL excluding the ES versions is a complete mess.

 

[CAT-THE-FIFTH]

Historically NVidia have been on the ball with OGL and Linux whereas AMD have been shocking. With Vulkan, AMD are now supporting that, so at least Linux has a chance. HL3 exclusive to SteamOS or Linux would have me dual booting and I am sure many others as well.

AMD is only supporting GCN1.2 cards ATM with Vulkan under Linux,unlike with windows where they support all GCN based cards!

 

[Mauller]

AMD is only supporting GCN1.2 cards ATM with Vulkan under Linux,unlike with windows where they support all GCN based cards!

That is only due to their newer Linux driver only supporting GCN 1,2 in the current implementation. Forgot the name of the system it is using.

 

[nashathedog]

I won't bother...

 

[Gregster]

Vulkan easier to work with?

It seems the hype train is running away on this one.

No, according to a dev from The Talos Principle, Vulkan is quite tricky to work with.

AMD is only supporting GCN1.2 cards ATM with Vulkan under Linux,unlike with windows where they support all GCN based cards!

I could have sworn I read that Vulkan was supported as far back as the 7 series cards unless I read it wrong (I did skim read).

-------------------------

Now encoding OpenGL Vs Vulkan in Vegas and will upload soon for those that are interested.

 

[humbug]

No, according to a dev from The Talos Principle, Vulkan is quite tricky to work with.

It will get better over time as they take feedback from devs and use it to make improvements.

 

[humbug]

AMD is only supporting GCN1.2 cards ATM with Vulkan under Linux,unlike with windows where they support all GCN based cards!

It ran on my 7870XT but not as well as the 290.

 

[CAT-THE-FIFTH]

I could have sworn I read that Vulkan was supported as far back as the 7 series cards unless I read it wrong (I did skim read).

It ran on my 7870XT but not as well as the 290.

I meant under Linux.

 

[r22snapper]

You obviously dont know the difference between offensive and informal. Look it up and dont ever call me a retard again.

Quite right. The term is "developmentally delayed."

 

[pete910]

AMD is only supporting GCN1.2 cards ATM with Vulkan under Linux,unlike with windows where they support all GCN based cards!

That's for the the new open AMDGPU driver that they have been working on. Their Cat driver is still there. Iirc the initial Vulkan release will be a closed "blob" anyhow.

 

[humbug]

I meant under Linux.

Are you sure its only GCN 1.2? the Fury series...?

GCN 1.0 only has 2 ACE Units compared with 8 on GCN 1.1 and 1.2.

Vulkan and DX12 are big on ASynchronous Compute, or at least its a big thing for AMD and the original GCN architecture isn't good for that.

I think GCN 1.0 is pre Mantle, it was developed along side GCN 1.1, GCN 1.1 incidently is what's in the XBox One.

GCN 1.1 = 290X/390X - 7790/260X