Valve: 'Vulkan is the Future, Not Much Reason to Create DX12 Backend'

[Dabombinable]

#Microsoft=rekt

[juretrn]

In the end, Microsoft will probably still prevail with DX12 just because they can just buy off devs to use it. 

[SirRoderick]

Is DX12 what console games will have to program for? If it is we could have a problem.

[Misanthrope]

Hey Valve: Fuck the future and give us something on the present like SteamOS already ffs: what's the fucking hold up it's been years, is not that hard to mod Debian you know

 

Otherwise the future will not happen.

[Davehaslanded]

I really couldn't care less what Low Level API is dominant. I just want to see some real world games showing off real world performance instead of constantly being told how great it is and being told how great the benchmarks are.

 

Like most users, I just want to be able to install a game and have it work well on my system. Do I want to see SteamOS succeed? Hell Yes!!  A free Gaming dedicated OS with the same low level benefits of consoles, but with the hardware flex-abilities of a PC is the holy grail and could see more console gamers jump ship. But steam have been telling us how big, great and game changing SteamOS and steam machines are going to be for ages now, and we still don't really have a system with what they were promising. Most of the first batch of steamboxes, ship with Windows or at least dual boot.

[ChineseChef]

-snip-

 

I very much agree with all of this.  I am tired of the hype, everything is hype train.  Maybe I have just become cynical, but I feel like every future tech is all hype and no substance.  The stuff either never comes to market and just kinds of gets forgotten, or it does show up but it isn't even half of what they claimed it would be.  Combine this with most of the markets becoming 1 massive company with a few small competitors that don't really compete, and it has become a race to maximize the status quo.

[escape093]

AMD's performance in OpenGL is pretty bad.

[werto165]

Can't switch to linux anyway, autodesk are not going to support it so I'm not going to move over. 

[Davehaslanded]

I really couldn't care less what Low Level API is dominant. I just want to see some real world games showing off real world performance instead of constantly being told how great it is and being told how great the benchmarks are.

 

Like most users, I just want to be able to install a game and have it work well on my system. Do I want to see SteamOS succeed? Hell Yes!!  A free Gaming dedicated OS with the same low level benefits of consoles, but with the hardware flex-abilities of a PC is the holy grail and could see more console gamers jump ship. But steam have been telling us how big, great and game changing SteamOS and steam machines are going to be for ages now, and we still don't really have a system with what they were promising. Most of the first batch of steamboxes, ship with Windows or at least dual boot.

 

I very much agree with all of this.  I am tired of the hype, everything is hype train.  Maybe I have just become cynical, but I feel like every future tech is all hype and no substance.  The stuff either never comes to market and just kinds of gets forgotten, or it does show up but it isn't even half of what they claimed it would be.  Combine this with most of the markets becoming 1 massive company with a few small competitors that don't really compete, and it has become a race to maximize the status quo.

So far the only thing Valve has really managed to do, is encourage OEMs to start building and selling ITX Windows gaming machines. I'm just tired from the 2-3 years of hype about all this and still no real action.

[Doobeedoo]

Would like to see Vulkan embraced by devs, would be amazing to see that.

[Sauron]

Money will win... if people don't learn to say to Microsoft "No" 

 

Microsoft won 20 years ago and has been sitting on the throne unchallenged since. These are just the preliminary steps to try and present a contender.

[ChineseChef]

So far the only thing Valve has really managed to do, is encourage OEMs to start building and selling ITX Windows gaming machines. I'm just tired from the 2-3 years of hype about all this and still no real action.

 

Valve's recent endeavors have become a really good example of "Perfection being the enemy of the good". 

[Litargirio]

I would love to switch to Linux and never ever use Windows again. I really hope that Vulkan becomes a thing.

[themaniac]

isn't Intel the ones developing Vulkan?

[deviant88]

Inb4 @ butthurt Micro$oft

 

" We are gamers "

" We will support games "

" Games for Windows Live "

                    ...

"Windows Gamers are XBOX Live customers" - Microsoft 2015

Get fucked ms.

[Davehaslanded]

isn't Intel the ones developing Vulkan?

Vulkan is being developed by The Khronos group which has a lot of developers and contributors. They can all be seen here: https://www.khronos.org/members/contributors

[Sintezza]

I really REALLY hope that the Devs take this advice to heart, and support Vulkan in a big way. Mainly because I REALLY do not want to switch to windows 10.

 

i totaly agree with this.

Windows 10 is just straight up garbage.

 

If game devs are realy smart, they would make their games to support both api´s.

This will be a big seling point.

Vulkan is indeed a great api, and the good thing about it is, that its platform undepended.

It can basicly run on anything.

[themaniac]

Vulkan is being developed by The Khronos group which has a lot of developers and contributors. They can all be seen here: https://www.khronos.org/members/contributors

ok thanks

[gluconolactone]

Why is Microsoft in the list of Khronos contributors? I really don't see how they can benefit from contributing in Vulkan's development.

[Humbug]

In the end, Microsoft will probably still prevail with DX12 just because they can just buy off devs to use it. 

They will co-exist. Neither API will kill the other.

DX12 for now will have more AAA games simply because it came out 6 months ahead of Vulkan.

But lots of devs will also get familiar with Vulkan because it is going to be on android, Linux Desktop and have support from all major non-EA game engines. Google's announcement that Vulkan will be the API of choice for android secures Vulkan's future. The only real question is will it be only Source2 games that use Vulkan on Windows, or will other games devs also try it on windows? It's similarity to DX12 means it will not be hard to move between them.

 

i think S2 was delayed partly because valve descided that they would only support opengl(vulkan) so they had ot go back and redo whatever they needed tooo

Source 2 was probably delayed because Valve.. :mellow: They have shipped it now anyway with OpenGL, DX11 and DX9 without waiting for Vulkan to go public, even though they do have it working on Vulkan internally. Also we can't really say it's delayed because they never announced a shipping date. Well now that they do finally have a game out with it and are fine tuning it the next question is when do they release the SDK tools so that other devs and content creators can start using it?

 

I'm just tired from the 2-3 years of hype about all this and still no real action.

Do you realise what Valve has achieved for Linux within a couple years?

Do you know that people now can and do play all Valve games on Linux including CS:GO, Left4Dead 2, Dota2, Half Life series, Portal series etc... Do you know that you can also play stuff like Shadow of Mordor, The Talos Principle, Witcher 2, The metro games, Bioshock infinite, the Civilization series, Dying Light, Ark Surivial, Rust, Kerbal Space Program, XCOM stuff, War Thunder, Company of Heroes 2, Serious Sam games, Postal games, Arma3 etc on Linux? In addition to over a thousand indie games a few of which are good (just like windows). Elite Dangerous for Linux comes out in December. Star Citizen is also coming out for Linux as is Project Cars, Ethan Carter, Arkham Knight, Grid 2. Ashes of the Singularity is coming out Linux using Vulkan. Alien Isolation for Linux comes out in a couple of weeks. Every major game engine with the exception of frostbite has now committed to supporting Linux which is going to make it easier for game devs to target Linux going forward. We also now that with Vulkan coming out this year Linux gamers will finally get DX12 level performance and reliability in AAA games and can say goodbye to openGL's terrible ecosystem. And Valve was a major factor in Vulkan development.

 

Yes Valve is a business, yes they are looking to secure their future, yes they always delay stuff, they do not release enough games and their steam support sucks etc... all these are valid points about Valve.

But let's not pretend that there is no action from them. They have dragged Linux gaming forward massively within a short time by canvassing other devs and supporting them to port to Linux as well as working on dev tools and behind the scenes API and ecosystem stuff. If you think there is no real action it's because you are comparing it with Windows PC, rather than against where linux gaming was 5 years ago.

 

They have also built this steam machine business model in such a way that (apart from Valve) none of the hardware makers are going to lose money on it if the 1st few iterations of Steam machines fails. They are made from PC parts and so manufacturing is very flexible and they can be sold off as normal gaming PCs etc and not much R&D tied up. The hardware makers are not taking a risk on this, it's not a disaster for them if they only sell small quantities. They can just iterate and come out with the 2nd gen and so on for a few years all the while growing a small niche market while simultaneously giving a boost for all the Linux desktop distros such as Ubuntu, Debian, Mint etc as more and more Linux games come out every month. So all these factors feed of each other's success and the overall Linux gaming ecosystem improves. It's different from the console business model where you need to sell loads of units in order to stay in business and you have a finite lifecycle.

 

btw I am very aware that Valve may indeed fail in the end, but if they do it will not be because of 'no real action'.

 

The future of Linux is mostly out of Valve's hands. It's upto the desktop Linux distros such as Debian, Fedora, Ubuntu, Mint and all the other guys to capitalize on Valve's work in gaming and push Linux forward as a viable consumer OS for PCs. That means that they will have to solve a bunch of other non-gaming related problems which people have with Linux which is outside of Valve's area.

[Davehaslanded]

Why is Microsoft in the list of Khronos contributors? I really don't see how they can benefit from contributing in Vulkan's development.

My guess is they wanted at least some input in an API designed to run on  their OS, especially as they move towards a more walled garden approach.

[Eroda]

snip

 

 

yeah i been waiting on S2 SDK thats what i have been wanting to see, a Source 2 tech demo similar vein to the 2003 source E3 One, 

[Davehaslanded]

-Snip-

I completely accept your points, all of which are totally valid, but a lot of what they are aiming for is still not actually here, right now. There is a lot of promise, and as i said, i'm desperate for them to succeed. If windows is treated like the unwanted child by AAA devs compared to the consoles, then Linux is the unwanted Nephew that they begrudgingly took on. There are very few releases for Linux that are simultaneous with their PC or Console counterparts. Could this change? Of course. If it doesn't change soon enough, could it prevent sales? Most definatley. Consoles have shown one thing. People will pick an eco system based purely on which one gets content first. 

 

I've got no real issue with their approach to the industry, but they started the fanfare way too early, and development has taken so long that people have become fatigued by it. (As i said originally this includes DX12) The other issue is the amount of times Valve have 'appeared' to backstep on a choice they made. Rightly or wrongly, it looks like weakness. (Xbone announcement ->launch anyone?) The Steam controller for example barely resemble the original prototypes. The steam-boxes 'feel' massively overdue (I know no release date was ever officially given) and by shipping most of them with windows, they've almost given up their USP.

[Humbug]

As for my thoughts, I think it's great for Valve to push for competition, pitting themselves against Microsoft and DX12. We just need to get some good performance numbers and I think we're all set.

When comparing different APIs the general public such as us likes to talk about performance numbers.

But for the devs what really matters is how good are the tools, is this easy to learn and work with, do I have a good ecosystem and debugging, documentation, are the drivers from GPU maker for this API solid and reliable and conformant? That's the stuff that's important, not the fps chart which we like to look at. Let's face it there is not going to be any real difference between a good DX12 implementation and a good Vulkan implementation.

e.g. if you put enough manhours into openGL even that can performs well but nobody likes doing that when the the API is a pain in the ass to work with and each GPU and each driver behaves differently.

[Briggsy]

Awesome, so valve are developing new games that use Vulkan, and will help pave the path forward for Vulkan? /s  /valvetime