DirecxtX 12 for ArmA 3?

[enex 11]

BrightCandle is it true that Dx12 is not something that makes performance better but simply ability to Make CPU talk

on multiple threads, which needs to programmed in a code of the game right?

[artisanal 22]

Dx12 isn't like a magic spell which automatically makes a game run with higher FPS. To access this kind of improvement, the developer needs to rethink/redesign how the engine works in some parts in order to get the best gains out of the new API. If this isn't done, there could be gains anyway, but not near as high as the things you can read here and there about dx 12.

Also, there's no need to emphasis the fact that every developer hates to fiddle around with his game engine for various reasons (mainly $$$). In conclusion, IF some day BI actually decides to port arma 3's engine to directx 12 (which would be really surprising) that wouldn't by far mean that you would get like +50% FPS.

Dynamic shadows would be nice tho' :bigglasses:

[brightcandle 114]

Dx12 isn't like a magic spell which automatically makes a game run with higher FPS. To access this kind of improvement, the developer needs to rethink/redesign how the engine works in some parts in order to get the best gains out of the new API. If this isn't done, there could be gains anyway, but not near as high as the things you can read here and there about dx 12.

Also, there's no need to emphasis the fact that every developer hates to fiddle around with his game engine for various reasons (mainly $$$). In conclusion, IF some day BI actually decides to port arma 3's engine to directx 12 (which would be really surprising) that wouldn't by far mean that you would get like +50% FPS.

Dynamic shadows would be nice tho' :bigglasses:

Right now the rendr process represents roughly 1/3 of the time of a frame. Assuming DX12 completely eliminated that whole section and it now ran instantly (not likely as pointed out the time taken is in BI code and not DX) then the end result would simply be to increase frame rate by a third. That is the absolute maximum DX12 can bring without fixes to other areas of the game loop. Of course with most of the time actually in BI code its going to be much less than that in practice.

[calin_banc 19]

NVIDIA and AMD Ready For Next Generation DirectX 12 API – Showcase New Features and Benefits of D3D12 API

Read more: http://wccftech.com/nvidia-amd-ready-generation-directx-12-api-showcase-features-benefits-d3d12-api/#ixzz3VBegEO9C

http://cdn3.wccftech.com/wp-content/uploads/2015/03/AMD-Command-Buffer-DirectX-11.jpg (238 kB)

http://cdn3.wccftech.com/wp-content/uploads/2015/03/AMD-Command-Buffer-DirectX-12.jpg (238 kB)

LE: What I find strange not to be an API limitation, is due to how those symptoms appear just as you increase the draw calls - happens in ALL RV engine games (ArmA 2, ToH, Day Z, ArmA 3, etc.), even on empty maps, devoid of any "simulation". On the other hand, DX11 might handle, let's say 10k draw calls on a powerful CPU, but for that to happen, the rendering part of an engine has to be optimized for such task extensively which in turn comes back again to poorly optimized said engine. Nevertheless, DX12 can bring significant improvement in visual elements that will effect gameplay and immersion.

Edited March 23, 2015 by calin_banc

[slim_pikins 12]

I do not think that for Arma3 we will see much more improvements with performance than we did with arma2 (which were massive btw when you look at the difference in the released game and the final iteration) DX12 isn't going to make a difference, but Arma 4 now that has the potential, the question for me is a business one 'is it financially viable to re-write the engine?' or 'is it financial suicide not to rebuild the engine?' Once people see what can be done with this 600% increase in graphics potential that developers are talking about are they going settle for less?

[DancZer 65]

Right now the rendr process represents roughly 1/3 of the time of a frame. Assuming DX12 completely eliminated that whole section and it now ran instantly (not likely as pointed out the time taken is in BI code and not DX) then the end result would simply be to increase frame rate by a third. That is the absolute maximum DX12 can bring without fixes to other areas of the game loop. Of course with most of the time actually in BI code its going to be much less than that in practice.

I have never did a deep profiling so i don't know whats are the limitations(and i'm not a game engine designer either), but i think it's kind a black box method without the source, isn't it?

How do you know:

-that the rendr process time is not limited by the engine to save ticks for the sim?

-that there aren't some pre/post parations for the rendr process to fit in that limit?(overhead)

The engine is built up to use single threaded DX(Arma 1, 2). Now its for multithreaded but single core, but maybe they didn't changed anything in the game engine design. Who knows? I think if they change a game engine design they can profit from DX12. Simply replacing the DX would make a minimal increase, because of the previous.

[calin_banc 19]

I do not think that for Arma3 we will see much more improvements with performance than we did with arma2 (which were massive btw when you look at the difference in the released game and the final iteration) DX12 isn't going to make a difference, but Arma 4 now that has the potential, the question for me is a business one 'is it financially viable to re-write the engine?' or 'is it financial suicide not to rebuild the engine?' Once people see what can be done with this 600% increase in graphics potential that developers are talking about are they going settle for less?

They can still "do their thing" if the community will continue to back them up, however getting new players will be a hard thing indeed. And then you have the problem of the competition to appear and take even more of your own player base.

[brightcandle 114]

I have never did a deep profiling so i don't know whats are the limitations(and i'm not a game engine designer either), but i think it's kind a black box method without the source, isn't it?

How do you know:

-that the rendr process time is not limited by the engine to save ticks for the sim?

-that there aren't some pre/post parations for the rendr process to fit in that limit?(overhead)

The engine is built up to use single threaded DX(Arma 1, 2). Now its for multithreaded but single core, but maybe they didn't changed anything in the game engine design. Who knows? I think if they change a game engine design they can profit from DX12. Simply replacing the DX would make a minimal increase, because of the previous.

Its not blackbox, because BI provide a debug build with a profiler in built we can actually see all the method names, all be it mangled a bit but much of the intent of them is clear for all the big pieces. So we know what is running and in what order and even the call structure of the code. We don't have the source code to work out why something is slow, but what is taking up time is easy to see. Your questions make no sense to a programmer and the spelling mistake in the second makes it even harder to fathom what it is you are trying to ask. There is no limit, the game is going as fast as it can and the evidence is in the profiler traces I have provided.

[anthariel 165]

https://twitter.com/maruksp/status/580026066183544832

Finally :yay:

[artisanal 22]

IF some day BI actually decides to port arma 3's engine to directx 12 (which would be really surprising)

Well, I'm really surprised now.

Dynamic shadows would be nice tho' :bigglasses:

[gibonez 18]

Awesome.

Looking forward to the increase in performance and how this might lead to larger multiplayer game modes .

[Fushko 60]

https://twitter.com/maruksp/status/580026066183544832

Finally :yay:

DX 12 CONFIRMED?! Not even waiting for Arma 4?!?

HURRY UP MS! :bounce3::bounce3::bounce3:

[chortles 263]

the question for me is a business one 'is it financially viable to re-write the engine?' or 'is it financial suicide not to rebuild the engine?'

This probably will have more to do with whether or not DX12-in-A4 happens than how good DX12 actually is. ;)

They can still "do their thing" if the community will continue to back them up, however getting new players will be a hard thing indeed. And then you have the problem of the competition to appear and take even more of your own player base.

Probably also a factor...

[slim_pikins 12]

That is my question answered :cool: in Arma 3, thank you BIS!

[enex 11]

This tweeter response means that Bohemia will work on low level stuff in arma 3? (as I understand

devs need to expend code under the hood regarding draw calls). I'm cautious : D

[DancZer 65]

Its not blackbox, because BI provide a debug build with a profiler in built we can actually see all the method names, all be it mangled a bit but much of the intent of them is clear for all the big pieces. So we know what is running and in what order and even the call structure of the code. We don't have the source code to work out why something is slow, but what is taking up time is easy to see. Your questions make no sense to a programmer and the spelling mistake in the second makes it even harder to fathom what it is you are trying to ask. There is no limit, the game is going as fast as it can and the evidence is in the profiler traces I have provided.

Sorry, but it's not my mother tongue.

That tweet is very promising.

[St. Jimmy 272]

What's wrong with Bohemia nowadays! Delivering bidops, ffv, afm etc. If DX12 does come and really improves performance then hats will fly around :dancered:

[Jackal326 1158]

What's wrong with Bohemia nowadays! Delivering bidops, ffv, afm etc. If DX12 does come and really improves performance then hats will fly around :dancered:

I was thinking more along the lines of pigs flying around, not hats :D

[chompster 29]

What's wrong with Bohemia nowadays! Delivering bidops, ffv, afm etc.

Maybe they finally got tired of people requesting it for the past 10 years. :icon_razz:

[slim_pikins 12]

I was thinking more along the lines of pigs flying around, not hats :D

Well I hope BIS can do it along these lines :)

[nikiforos 450]

That will be the only reason for me to upgrade to Win10( if we actually can see improvement in performance) I just hope that my GTX 780 will be good enough to take advantage of DX12.

[forteh 11]

That will be the only reason for me to upgrade to Win10( if we actually can see improvement in performance) I just hope that my GTX 780 will be good enough to take advantage of DX12.

From what I've read, all nvidia cards going back to the 400 series will support dx12. I will probably upgrade regardless of arma3 performance increases, if it does help then great but I'll not be devastated when it's not the magic bullet :)

[calin_banc 19]

http://community.amd.com/community/amd-blogs/amd-gaming/blog/2015/03/26/amd-enables-incredible-directx-12-performance-in-new-3dmark-api-overhead-feature-test

Perhaps those with weaker CPUs should enjoy a nice bump in performance when DX12 hits ArmA 3!

.

.

.

.

.

.

At least in some cases. :D

[I give up 152]

To me, this DX12 hype looks like a elaborated marketing campaign.

After Windows 8 flaw, Microsoft had to do something to lead consumers to get Windows 10, otherwise Windows 10 would have the same destination of Windows 8. No one is going to drop Windows 7 for a matter of games, even more when Windows 10 requires a huge loss of privacy.

DX12 is the way to convince gamers for a update and also is good for hardware manufacturers, lots of people will bury hardware just to get the newest awesome DX12 ready.

And when AMD is leading the hype for DX12, it makes me even more suspicious, can't forget when DX11 came out the troubles that AMD cards had, to make cards usable they had to add a option to disable Tessellation.

Let's see.

[Dwarden 1124]

Bratwurse, everyone knows it, DX12 is the main foundation to sell new Windows, 8.0/8.1 kernel was quite good btw. (UI was disaster)