DirectX 12: The future of it within the console gaming space (specifically the XB1)

I seen many throw this assertion out there. And it seems to be a rather easy narrative to paint especially when no one is talking (MS, AMD or Nvidia). But one thing seems odd to me.

Doesn't MS producing a low level api especially a general api compatible with a broad swath of hardware require cooperation from both NVidia and AMD?

Seems like to me Nvidia dragging its feet would be the more likely source of the delay of the development and release of DX 12.

Mantle should have scared Nvidia far more than it should have scared MS.
We saw many times how bad and arrogant MS managment is recently. Windows 8, xbone. They did not think about it at all. Only when they got a lot of flak they have started doing things.
 
We saw many times how bad and arrogant MS managment is recently. Windows 8, xbone. They did not think about it at all. Only when they got a lot of flak they have started doing things.
again, you don't plan and write code for e.g. DX12 in 2 years. I think you don't know how long software cycles can be. Yes in the last two years, you can be sure there were some adjustments, just to react on new situations (like mantle) but something like that does not appear over night. When something get's official (two years ago), you can bet they are working at least one or two years on it.
And MS wasn't arrogant with win 8 or with the xbone. They just thought in other ways. They thought it would be great if you get a all-in one media, gaming, social (e.g. Skype) and online device. And xbox is doing well. Yes PS4 is doing event better, but that doesn't change the fact that it is also doing well. they just don't wanted to oversleep another trend ... that's all.

offtopic:
btw. windows 8 is a really great OS for newcomers, e.g. my parents had no problem with it. And for those who don't want the new GUI stuff, just install e.g. classic shell and get rid of it.
Also, it is a great OS that is still compatible with really really old software (win 95). Not all of it, but most. That alone is a giant achievement and a giant burden.
 
Oh they were arrogant very much.
With Win8 they shoved metro interface unsuitable for productive work on desktop.
With xbone see all DRM fiasco and "180"s.
 
If you just want an anti-MS rant, NeoGAF would be a more appropriate forum. I believe etiquette here is to backup your assertions with evidence.
They'd launch it sooner on xbox one, if they did. Xbox launch SDK was abysmal.

With W8 and xbox one launch you can see how confidently arrogant, delusional MS was.
 
They'd launch it sooner on xbox one, if they did. Xbox launch SDK was abysmal.

With W8 and xbox one launch you can see how confidently arrogant, delusional MS was.
Can you start supplying a little more support for views rather than just hyperbolic vitriol? Your current phrasing doesn't aid discussion any and if you've nothing more to add beyond this POV, you should leave it there.
 
Yes, in the light of what we know about the XboxOne and DirectX12 on PC, those figures make a whole lot more sense.
 
Just an update on the PCars. 30-40% for PC. 7% or more on Xbox One. Hmmm I'm going to read between the lines for MJPs post now lol. I'm not sure what dx12 is providing though over dx11.X. Would be an interesting discussion to find out what.

http://forum.projectcarsgame.com/sh...ame-rate-issue&p=952697&viewfull=1#post952697

30 to 40% were the average gains ON PC with games like Battlefield 4 and Thief using Mantle. So it was expected that these values were on PC.

7% on Xbox may seem very little in comparison, but I guess the CPU usage was already manually optimized so there was little to be gained. And since D3D overhead is already near zero with current API...
 
30 to 40% were the average gains ON PC with games like Battlefield 4 and Thief using Mantle. So it was expected that these values were on PC.

7% on Xbox may seem very little in comparison, but I guess the CPU usage was already manually optimized so there was little to be gained. And since D3D overhead is already near zero with current API...
Well you need to jump over to the DF thread where MJP makes a post about async copy. Anyway, it's not important. I didn't see Xbox gaining 30-40% by moving to 12. DX11.X does contain most of the features of 12, I'm still just trying to figure out which ones its missing. This particular comment (from PCars dev) helps shed some light on what they've been doing with the current version of the game, if he is to be believed however, it gets confusing with as 'the main' benefits from D3D12 should be:

a) low overhead - check + 1, xbox extensions go deeper than 12.
b) Multi-engine (3d, copy, compute) - check
c) multi-adapter
d) multi-threaded command buffer submission - ?
e) hardware level features - check
f) resource binding - check + 1

maybe I'm missing some other major items, this is the best that I can do.
 
What really needs DX12/Mantle is mobile! Draw calls kill it, as I've just discovered with my simple 2D game. Up to a little over 100 calls on a Galaxy S3 Mini phone (800 MHz Cortex A9) just drawing sprites and it can't manage 60 fps. The final game will be more efficient after optimisation, but draw calls are really holding mobile back (also talked about a lot among mobile devs that I've seen). That's another area Windows could have a noteworthy advantage on mobile, I guess. Not sure Vulkan will gain speedy adoption.

Impact of DX12 on consoles is thus a significant advance for Windows microconsoles, I think. Would love to see some low-power (Atom) Windows devices benchmarked.
 
Vulkan should end up in Android at some point, no? Then it would just be a matter of a couple of years for it to be pervasive. Apple has Metal and Windows phone has DX12, so the push for low-level optimization should be real.
 
Vulkan should end up in Android at some point, no? Then it would just be a matter of a couple of years for it to be pervasive. Apple has Metal and Windows phone has DX12, so the push for low-level optimization should be real.
Android should first be quicker to use itself -I have an Androind phone and I am tired of it, and how it bad it is managing battery life- and then it could go to the metal.

the mustard race (to say something to call smartphones something) is not the kind of race that cares about optimisation and intelligent use of resources. I'd be disappoint if say..., sebbbi's boss put him working on smartphones.
 
I have spent a week optimizing a 40 instruction (ALU bound) shader, and in the end I got rid of a single instruction from it (2.25% gain for 2 millisecond period). It was worth it. Console graphics programming has never been about what you can do. That 16.6 millisecond time slice tells you what you can do, you just try to cram in as much as possible and cut out the rest.
If that's a pixel shader, wouldn't 40 ALU cycles (if indeed it's 40 cycles) be fillrate limited?
 
IOS did already introduce Metal API last year if you remember, but sebbbi I think pointed out some serious limitations that are holding it back currently from truly performing. Who knows though with the next iOS release they will fix that.
 
This guy also says: "It gives us more control over multi-threading which in turn has less CPU overhead. Our engine is massively multi-threaded.

The render team can talk more on the technicalities as it's rocket science to me."

Assuming that he is right though for the sake of it, then it would confirm PCars runs GNMX, which is Sony's path for easy DirectX ports ...

Mr Rocket science is now saying that 30% is for the PC and the XB1 might see 7%.

It was obvious to most, but the secret sauce group was holding firm.
 
Mr Rocket science is now saying that 30% is for the PC and the XB1 might see 7%.

It was obvious to most, but the secret sauce group was holding firm.
Its not even 30-40% on PC. Thats in the best case scenario for people with old quadcore system. In reality its anywhere from single digit to a max of 30-40% improvement. If you have a graphics settings that don't push your gpu (eg low settings) or have a powerful gpu, coupled with a low clocked quadcore you would have been drawcall bottlenecked on dx11, on dx12 there would be huge performance improvements as multiple slow clocked cpu cores can issue drawcalls instead of just one core.
 
Back
Top