Vulkan 1.0 Graphics API Brings Cross-Platform Performance Boost; Intel, AMD, Nvidia Contribute

Status
Not open for further replies.

kenjitamura

Distinguished
Jan 3, 2012
195
3
18,695
Vulkan is completely open-source
Small nitpick here but that's wrong. Vulkan isn't open source it is an open standard. To make changes to the Vulkan specification itself, that will be universally adopted by all other Vulkan adopters, you will need to be a Khronos member which is something like $10,000 a year for companies but less for universities.

There will be open-source implementations of Vulkan that anyone can contribute to for free but any big changes like new extensions won't be universally adopted by all implementations of Vulkan in places like game engines so it won't really affect Vulkan as a standard.
 

xHDx

Honorable
Nov 11, 2012
55
0
10,630
It's all nice and all but there is still uncertainty.

Firstly, Vulkan is probably at least 40%+ Mantle. If you followed Mantle, then you know this is the truth. There are many articles online explaining the story.
Secondly, there is the problem of this not taking off. Maybe it does offer more than what DirectX does or OpenGL, but I doubt many will adopt it. Mantle didn't take off, so I can't see this being getting much further. Maybe im wrong. I hope not, but who knows.
 

Bloob

Distinguished
Feb 8, 2012
632
0
18,980
It's all nice and all but there is still uncertainty.

Firstly, Vulkan is probably at least 40%+ Mantle. If you followed Mantle, then you know this is the truth. There are many articles online explaining the story.
Secondly, there is the problem of this not taking off. Maybe it does offer more than what DirectX does or OpenGL, but I doubt many will adopt it. Mantle didn't take off, so I can't see this being getting much further. Maybe im wrong. I hope not, but who knows.

There are drivers already out, game engine -vendors are already implementing a Vulkan version of their rendering pipeline. It's not sure to take off, but with "global" support in desktops and mobile, it seems quite likely.
 


Mantle didn't 'take off' because nVidia refused to support it, and because AMD kept it for themselves since it was still in beta. They were trying to figure out whether to make it open as is, or doing something else to increase adoption. Rather than opening it up, they gave Mantle to Khronos instead, since they knew it would not get support if nVidia refused to adopt it.

Vulkan will very likely be successful. Since it's on a lot of different platforms, it will make porting easier for developers.There's little reason not to adopt it. Aside from that, it works even on Windows XP. Lots of people are refusing to upgrade from Windows 7 to Windows 10. DX12 is limited to W10, while vulkan works across pretty much all windows platforms... That alone increases the potential buyers for developers over DX12, so... It should do fine.
 

eklipz330

Distinguished
Jul 7, 2008
3,033
19
20,795
I foresee massive backpedaling by Microsoft if those shows even an ounce of adoption over their API... Maybe win 8 support, maybe lower or no royaltied... Competition is good.
 

Djibrille

Distinguished
Jan 31, 2007
27
0
18,530
"Additionally, Vulkan is actually able to handle multiple CPU threads effectively, which can resolve the CPU bottlenecks from which many OpenGL applications."

I think this phrase "suffers" from a missing word(or more?).
 

hannibal

Distinguished
The diversity of vulkan is the key. The problem is, as mentioned in the article, that it is harder to program.
But if you can get DX12 level graphic in windows7 should make this API appealing to many companies. On the other hand they can chose to only use DX12 and force everyone to win10... But there definitely is potential in Vulcan to become very popular api!
 

hst101rox

Reputable
Aug 28, 2014
436
0
4,810
This should give more respect to the intellectual abilities of the klingon people. For too long, they have been considered merely great warriors with guts and glory with hardly the ability to develop complex systems without special drugs to inhibit their, sometimes considered, barbaric behavior to other races, now banned.
 
Vulkan, mantle, dx12 - I just want to see something viable, actually in use to a consumer. Something more concrete than unicorn farts which is so far what all 3 of these have produced. A lot of hot air, not better gaming on my screen. It's like everyone huddles around talking about and debating the dream the other one had last night and in the end - it doesn't matter because it never actually happened.
 

bit_user

Polypheme
Ambassador
I expect a wave of open source libraries will ease the burden on programmers - especially those who don't need access to all the low-level details.

But lots of games are built atop a small number of engines, the developers of which have been clamoring for the Mantle/Vulkan/DX12 capabilities. So, those games will get a virtually "free" benefit, as the engines are updated to use Vulkan.
 

Myrmidonas

Distinguished
Mar 13, 2013
128
3
18,715
Vulkan, mantle, dx12 - I just want to see something viable, actually in use to a consumer. Something more concrete than unicorn farts which is so far what all 3 of these have produced. A lot of hot air, not better gaming on my screen. It's like everyone huddles around talking about and debating the dream the other one had last night and in the end - it doesn't matter because it never actually happened.

I couldn't have said it better. Very well put.
 

bit_user

Polypheme
Ambassador
Mantle was AMD-specific, and they didn't even let the general developer community use it until some time after DX12 and Vulkan were already announced, IIRC. Understandably, few developers made the investment in supporting it, though some did.

DX12 requires the latest hardware and Windows 10, so games are understandably slow top adopt it (plus, rewriting your backend takes time).

And Vulkan is an open standard, which must be negotiated by all the stakeholders (game developers, mobile & desktop GPU folks, and OS folks), and go through all the hurdles of Kronos' standards process, which takes time. has only just been finalized, yet there's already some driver support and games.

I think you need to keep this type of news in context, and have realistic expectations for when this stuff will begin hitting your screen.
 

ZolaIII

Distinguished
Sep 26, 2013
178
0
18,690
As a member of Croteam said it's not so hard to program. Structure is more complex & it do needs some additional time in development process to be done which naturally costs but compared to the time, money and effort needed to port a same game between different apis on different platforms it's dirty cheap. As you forgot to mention one significant advantage of Vulcan I will mention it for you. Making a GPU driver for Vulcan is much easier & less complex than for OGL & we all know how most of GPU maker's never had a decent OGL driver's. The sheder structural management (that can save a lot of RAM) & easier to make GPU driver's alone more than compensate for additional structural complexity.
 
I know why certain technologies are taking awhile to reach the end user but to be fair, mantle was being hyped up in tech magazines well over a year ago. Where is it? That's my point. First it was mantle, then it was vulkan, mantle got sort of shelved along with vulkan. It explains why vulkan was delayed with so many companies involved. Mantle sort of quieted down with microsoft announced dx12 and even amd was pushing support for dx12.

In the meantime we've had close to 2yrs of musical chairs regarding 'the next best thing' for graphics api's and yet we're all still gaming on dx11. Nothing new. Only a couple of demo's and maybe one beta game making much use of any of it. I don't care if they side with android and call it chocolate syrup and nuts, however it would be nice to see anything materialize at this point.
 

heffeque

Distinguished
Dec 9, 2002
181
0
18,680
Well, taking into consideration that Vulkan is an open evolution of Mantel, and some companies have already been "messing" with Mantel, we'll have Vulkan support sooner than it took to have Mantel support. Heck, even nVidia has already outed a beta driver supporting Vulkan.
 

tamalero

Distinguished
Oct 25, 2006
1,125
133
19,470
Vulkan is completely open-source
Small nitpick here but that's wrong. Vulkan isn't open source it is an open standard. To make changes to the Vulkan specification itself, that will be universally adopted by all other Vulkan adopters, you will need to be a Khronos member which is something like $10,000 a year for companies but less for universities.

There will be open-source implementations of Vulkan that anyone can contribute to for free but any big changes like new extensions won't be universally adopted by all implementations of Vulkan in places like game engines so it won't really affect Vulkan as a standard.
Pretty sure they mean FREE as in using and programming on it (like PHP), but doesnt means you can modify the standards just because you want.
 

Ra_V_en

Honorable
Jan 17, 2014
1,296
0
11,960
Pretty straightforward conclusion... give me Star Citizen with Vulcan support and I'll finally will be able to trow away M$ OS forever, thank you.
I've foreseen this coming just after the Vulcan appeared in the news and the more it comes reality more smile on my face. Linux for entertainment, Windozz for office crapware I can live with that no problem. Wondering what M$ is going to do now, more subscriptions on enterprise software?
 
Things take TIME in the software world.

Mantle as said is turning into Vulkan. There is a REASON that we need an API like Vulkan and that is for scenarios where DX12 won't work.

- SteamOS
- other Linux (possibly PS4 later)
- mobile
- other hardware that DX12 won't support.

As said it takes TIME, but Vulkan and the game engines, tools and extended ecosystem will continue to improve.

Making a game will continue to get easier to the point that we MAY see fairly easy development of certain games to mobile, W10, SteamOS, PS4 and XB1 all at the same time (most of the core code at least).

I don't know if Microsoft can force DX on the XB1 (they do want to have XB1, W10 PC and mobile W10 devices in their own ecosystem), however it's possible Sony may allow Vulkan once things improve.

I'm not sure if there's any financial incentive for Sony to discourage Vulkan. If the overall cost is lowered (they develop the PS4 API, game integration etc) and the reliability and performance is as good or better I suspect they'll just switch over.

That would be HUGE, and Microsoft would be force to justify DX on their ecosystem with the obvious ramifications being if you don't need Windows then how many would go with Linux instead (especially with SteamOS).

*On a side note, people speak about the "failure" of SteamOS and I think that it will be a huge success. However, like Vulkan, these things take time. I think the bigger question is:

What would it take for Vulkan and/or SteamOS to be failures?
 
Things take TIME in the software world.

Mantle as said is turning into Vulkan. There is a REASON that we need an API like Vulkan and that is for scenarios where DX12 won't work.

- SteamOS
- other Linux (possibly PS4 later)
- mobile
- other hardware that DX12 won't support.

As said it takes TIME, but Vulkan and the game engines, tools and extended ecosystem will continue to improve.

Making a game will continue to get easier to the point that we MAY see fairly easy development of certain games to mobile, W10, SteamOS, PS4 and XB1 all at the same time (most of the core code at least).

I don't know if Microsoft can force DX on the XB1 (they do want to have XB1, W10 PC and mobile W10 devices in their own ecosystem), however it's possible Sony may allow Vulkan once things improve.

I'm not sure if there's any financial incentive for Sony to discourage Vulkan. If the overall cost is lowered (they develop the PS4 API, game integration etc) and the reliability and performance is as good or better I suspect they'll just switch over.

That would be HUGE, and Microsoft would be force to justify DX on their ecosystem with the obvious ramifications being if you don't need Windows then how many would go with Linux instead (especially with SteamOS).

*On a side note, people speak about the "failure" of SteamOS and I think that it will be a huge success. However, like Vulkan, these things take time. I think the bigger question is:

What would it take for Vulkan and/or SteamOS to be failures?
 
Status
Not open for further replies.