DirectX and OpenGL to get low level access according to leaked statements (GDC 2014)

[TerryMathews]

Its a true statement, is it not?

 

[Carfax83]

There's no way(that I see)that MS can possibly build a new API with Mantle's granularity because they do not design or build the hardware.

Microsoft does not build hardware, but they obviously have indirect input in the design because they create software that necessitates hardware support.

Even if they do release a lower level DX edition it will still have to use some kind of abstraction layer to interface with the video card(and CPU if AMD)and therefore have less ability to fully exploit the architectures of either,compared to Mantle.

So? Mantle uses abstraction as well, and even the consoles do. Close to metal is a hype phrase for the most part..

NV will release their own(Mantle) which will serve a similar purpose but still require abstraction.

An opinion which has not one single shred of supporting evidence.

 

[3DVagabond]

Mantle has more market penetration than DX12 does

Probably has more than Win 8.1 as well.

 

[Gloomy]

Mantle isn't competing with DX. It's just an AMD feature they can get into games similar to Nvidia's exclusive PhysX effects.

Whatever Microsoft is doing with a future DX release that will have parity with Mantle (making Mantle's business case disappear) is years away still. And when it's released, it will still have less market penetration than Mantle. So AMD has plenty of time to get Mantle into games.

I don't see why Mantle needs to "take out" anything. It clearly performs better in BF4 than DX (VRAM management aside, as that will get fixed eventually), as such it presents a selling point for their cards. And it will presumably do so in every game it's built into.

 

[3DVagabond]

Both financially, and practically. Changing driver models, especially to the degree that was done between XP and Vista, is not trivial I'm sure, and would have required a major rewrite of the OS's code.

In instances like that, I think it's better to start from scratch and do it properly for the greatest impact. Microsoft has shown that it's willing to break compatibility between DirectX versions to move the technology forward, and DirectX sometimes requires a new driver model to add or expose new features.

DX9 to DX10 was such a paradigm shift..

That's ancient history. What about DX multi threading in Win 7? Seems to be working great in Win 8.1, but not Win 7.

Microsoft does not build hardware, but they obviously have indirect input in the design because they create software that necessitates hardware support.

With Microsoft controlling DX and it's feature set, we will likely never see any real differentiation between nVidia and AMD feature sets as Microsoft sets the feature level. I'm sure they have input from the hardware manufacturers, but in the end if Microsoft decides a certain feature is in the too hard pile, or even if AMD or nVidia scream and holler too much certain features will get left out. I'm hoping that all of them getting together for this will mean that each manufacturer's hardware will get it's own support for specific functionality.

 

[Gloomy]

AMD sez:

AMD would like you to know that it supports and celebrates a direction for game development that is aligned with AMD’s vision of lower-level, ‘closer to the metal’ graphics APIs for PC gaming. While industry experts expect this to take some time, developers can immediately leverage efficient API design using Mantle, and AMD is very excited to share the future of our own API with developers at this year’s Game Developers Conference.

http://www.pcper.com/news/Graphics-Cards/DirectX-12-and-new-OpenGL-challenge-AMD-Mantle-coming-GDC

"Use Mantle now, because that's where DirectX is heading."??

 

[Spjut]

It seems like DX10 on XP hadn't been an easy task. Read this for example
http://www.beyond3d.com/content/articles/55/2

But Microsoft did backport WDDM 1.1 and DX11 to Vista. I don't see why it would be so hard backporting the improved driver models to Windows 7 (or Vista too for that matter)

No matter how much DX11.3 or DX12 is improved, very few studios will use it unless it comes to Windows 7. It'll be just like XP and DX9, DX10 had for sure gotten adopted much earlier if it didn't mean abandoning the XP userbase

 

[OCGuy]

AMD sez:



http://www.pcper.com/news/Graphics-Cards/DirectX-12-and-new-OpenGL-challenge-AMD-Mantle-coming-GDC

"Use Mantle now, because that's where DirectX is heading."??

Have any developers jumped on board since the BF4 fiasco/thief Mantle delay?

Will we ever see a developer use Mantle outside of the Gaming Evolved program?

The burden of proof lies on the person claiming something outside of the status quo.

 

[mikk]

We’ve spoken to several sources with additional information on the topic who have told us that Microsoft’s interest in developing a new API is a recent phenomenon, and that the new DirectX (likely DirectX 12) will substantially duplicate the capabilities of AMD’s Mantle. The two APIs won’t be identical — Microsoft is doing its own implementation — but the end result, for consumers, should be the same: lower CPU overhead and better scaling in modern titles.

http://www.extremetech.com/gaming/1...rectx-12-will-imitate-and-destroy-amds-mantle

 

[ShintaiDK]

Have any developers jumped on board since the BF4 fiasco/thief Mantle delay?

Will we ever see a developer use Mantle outside of the Gaming Evolved program?

The burden of proof lies on the person claiming something outside of the status quo.

Seems everyone that could ran away.

http://www.amd.com/us/products/technologies/mantle/Pages/mantle.aspx#2

I remember people claiming basicly all FrostBite3 games would have Mantle. Now it seems only 1 game will. 3 Mantle games for 2014, and maybe another or 2 for 2015?

But again, who wants to pour money in something that at best can only be for 18% of the market. 35% if we look on dGPUs.

 

[OCGuy]

http://www.extremetech.com/gaming/1...rectx-12-will-imitate-and-destroy-amds-mantle

http://n4g.com/news/1362360/rumor-amds-mantle-api-is-a-direct-copy-of-xbox-ones-low-level-api

AMD recently announced their new Mantle API. This API has been said to give AMD a significant advantage over their competitor Nvidia in terms of in-game performance. Anandtech speculates that AMD might be confident of this because their Mantle API is a direct copy of Xbox One’s Graphics API allowing developers to directly port their games from Xbox One to PC. Since Mantle can use Direct3D high level shader language, Xbox One’s shader will be already coded in the first place so there is not much work involved in porting them to PC

 

[OCGuy]

I remember people claiming basicly all FrostBite3 games would have Mantle. Now it seems only 1 game will. 3 Mantle games for 2014, and maybe another for 2015?

People claimed a lot of things in that infamous Mantle thread :biggrin:

Reality is that FB3 is said to be "Mantle ready". Somehow that started a rumor that every game with engine automatically uses the Mantle API.

 

[cytg111]

A quick google on windows versions, d3d and opengl .. you know, instead of making my own stuff up.

http://www.extremetech.com/gaming/133824-valve-opengl-is-faster-than-directx-even-on-windows

With an Nvidia GTX 680, Intel i7-3930k, and 32GB of RAM, Windows 7 and DirectX, Left 4 Dead 2 maxes out at 270.6 fps. With the same hardware, but different software — Ubuntu 12.04 and OpenGL — L4D2 scores 315 fps, almost 20% faster than Windows.

Yea, microsoft is sure rocking that boat, as they allways do they attain monopoly on a segment... oh wait!

http://www.techpowerup.com/118477/amd-first-with-opengl-4-0-compliant-graphics-driver.html

OpenGL 4.0 is comparable and up to times with Microsoft's DirectX 11 API, it makes use of hardware features such as tessellation on the GPU, per-sample fragment shaders and programmable fragment shader input positions, 64-bit double precision floating point shader operations, etc., and has no restrictions on which later version of Windows it can run on. With OpenGL 4.0 for example,one can expect 3D graphics with the complexity comparable to DirectX 11 on Windows XP. "

Yes, we obviosly need a new operating system and a new drivermodel to unlock the heavenly goods of .. of what opengl does on windowsxp ... Yup, nothing to see here folks, please move along. *cough* * *cough*.
That is microsoft, as we know it, hard at work... giving me an incentive to upgrade my operating system just to get access to the next level of directx

 

[SiliconWars]

You know if DX12 basically copies Mantle's multi-threading - and why wouldn't it - it's just going to make Mantle even easier to implement for devs.

There's nothing much stopping Nvidia cards from running on Mantle already remember. They just wouldn't get the GCN-specific benefits - and that's what AMD would use Mantle for in the event of a DX12 with Mantle-influenced multi-threading.

 

[Warsam71]

Hello everyone,

Here is the official statement we have released about this topic:

[FONT=&quot]"Yesterday several articles were published that reported that DirectX and OpenGL are being extended to include closer-to-metal functionality and reduced CPU overhead. AMD supports and celebrates a direction for game development that is aligned with AMD’s vision of lower-level, ‘closer to the metal’ graphics APIs for PC gaming. While industry experts expect this to take some time, developers can immediately leverage efficient API design using Mantle, and AMD is very excited to share the future of our own API with developers at this year’s Game Developers Conference."


[/FONT]

 

[MutantGith]

You know if DX12 basically copies Mantle's multi-threading - and why wouldn't it - it's just going to make Mantle even easier to implement for devs.

There's nothing much stopping Nvidia cards from running on Mantle already remember. They just wouldn't get the GCN-specific benefits - and that's what AMD would use Mantle for in the event of a DX12 with Mantle-influenced multi-threading.

I'm always hearing this bandied about.
Where is the document containing the minimum feature set for the Mantle API? Failing that, where has someone from AMD actually said that NVIDIA cards can run the Mantle API, and detail a mechanism by which NVIDIA could pursue this goal?


Since people are speaking from a position of fact, there must be documentation about this out there. If that's the case, then we can see which generations of which graphics cards work within that minimum feature set. While I agree that the comments from developers seems to indicate that Mantle is far more about programming tools and good practices baked into the API, as far as I know there hasn't been a direct statement as to which hardware it would work for. If it is so easy to be supported on NVIDIA chips, why hasn't support been rolled out for 6 series and 7 series AMD cards?

And even if you were right, and it would be doable on NVIDIA, and DX12 brings the same thread scaling benefits as Mantle, Why would developers support Mantle in addition to the DX12 path that you surmise would show all the same benefits? I find it hard to believe that people would up their coding, QA, and development budget significantly, just to support a redundant architecture for the slim number of people who (at that point) would be running Win 8.0 or maybe 8.1 (depending on the DX 12 supposed roll out timeline) AND a post 7900 GCN card.

Again, maybe my neuroticism, but is seems like there is just waaaay too much inference being thrown about. It seems somewhat less likely that we'll end up with a fractured, 1 API per vendor computing future based on this news. That seems like a win for everyone.

 

[Mand]

I'm actually more surprised to see an AMD guy there given it completely undermines their whole Mantle strategy.

That assumes that their Mantle strategy isn't to cause exactly this sort of development, which to me seems entirely possible.

 

[SiliconWars]

I'm always hearing this bandied about.
Where is the document containing the minimum feature set for the Mantle API? Failing that, where has someone from AMD actually said that NVIDIA cards can run the Mantle API, and detail a mechanism by which NVIDIA could pursue this goal?

Well considering the sdk hasn't been released yet and it's still beta software, it's asking a bit much. I don't see why AMD should detail how Nvidia can make use of it anyway?

All of the devs who have access to it so far have said that there is nothing inherently preventing Nvidia from using Mantle though. It's not hard-wired to GCN. Why not believe them?

If it is so easy to be supported on NVIDIA chips, why hasn't support been rolled out for 6 series and 7 series AMD cards?

Because it's still in beta? Even at that it's barely just in beta.

And even if you were right, and it would be doable on NVIDIA, and DX12 brings the same thread scaling benefits as Mantle, Why would developers support Mantle in addition to the DX12 path that you surmise would show all the same benefits? I find it hard to believe that people would up their coding, QA, and development budget significantly, just to support a redundant architecture for the slim number of people who (at that point) would be running Win 8.0 or maybe 8.1 (depending on the DX 12 supposed roll out timeline) AND a post 7900 GCN card.

You have evidence of all that? Because anything I've read has said the opposite - ie Mantle is *not* difficult to develop for. If the multi-threading was a basic copy/paste job from DX12 then that would be half the job done already.

 

[MutantGith]

Well considering the sdk hasn't been released yet and it's still beta software, it's asking a bit much. I don't see why AMD should detail how Nvidia can make use of it anyway?

All of the devs who have access to it so far have said that there is nothing inherently preventing Nvidia from using Mantle though. It's not hard-wired to GCN. Why not believe them?

Because that is a vague marketing slide. Because they haven't filled in any details, despite praising this API as being revolutionary for months. Again, the issue comes down to consistency and confusion in messaging.

Initially, A PR rep from AMD did say that Mantle consists of a GCN card as a necessary part of the equation. I'm not going to go look it up, but it was in that horrific locked Mantle thread. It took place at some time relatively early in the release process, and so you could make the claim that it has been invalidated by the conference wherein they presented the slide you've included above.

There's still a logical issue with that though.

This slide you've copied again shows that the API "does require a minimum feature level from hardware". Great. That's incredibly vague.Why not, in the intervening months, now that the second game using this API is set to sometime soon get a post-release patch fully describe the requirements?Currently they are saying that Mantle is not necessarily hardwired to GCN...forever. However, functionally, AMD could mandate a minimum feature set that effectively excludes everything but GCN cards, and be technically correct and clear on that statement.

Effectively, they aren't lying if NVIDIA and Intel could use the API in the future, as long as NVIDIA and Intel exactly replicated all of the function of a current GCN card, either though abstraction, emulation, or paying big licensing fees to AMD. Therefore, in this scenario, Mantle is not actually 'hardwired' to GCN, just so much trouble for everyone else to try and use that it is effectively only for AMD hardware.

I'm not saying this is the case, but it smacks a little bit of mixed messaging when AMD proudly claims that anyone can use it, and then obfuscate exactly what it really takes to go ahead and do that.

If it is inclusive, and can run on other hardware, why not just publish the spec? Let people know what the specifications are, and how one goes about trying to program for it.
It would seem if AMD were trying to convince people how easy and inclusive Mantle is, showing them how easy and inclusive it is, IE showing how easy it is to have hardware conform to that minimum feature set, would be a good move.

Weather it is in beta or not, AMD should know the baseline requirements for their own API. I would hope. If so, why not publish or at least release them>? I'm inclined to think that it's because they haven't even been able to fully roll out full featured support for their own GCN 1.0 cards yet, and to try and pitch other IHV's using it while they are having issues would seem silly.

You have evidence of all that? Because anything I've read has said the opposite - ie Mantle is *not* difficult to develop for. If the multi-threading was a basic copy/paste job from DX12 then that would be half the job done already.

Again, the point seems to have been lost. In your example, if half of the job is done already from cut and paste from DX 12, you see that as a win. Looking at it the other way, the other half, the half that isn't done still has to be budgeted for. Again, two scenarios.

1) Code for the newer, slimmer DX implementation, that works for everyone with capability reporting scaling, from Dx 9 cards up, from all three manufacturers. Perform QA, code support, and pipeline new games as has been normal practice for years, using DX.

or

2) Code for the newer, slimmer DX implementation, that works for everyone with capability reporting scaling, from Dx 9 cards up, from all three manufacturers. Perform QA, code support, and pipeline new games as has been normal practice for years, using DX. Copy and paste half of the code from the DX version into the Mantle version, troubleshoot the problems caused thereby, and now code the remaining Mantle implementation, troubleshoot that code path, and maintain separate QA and programming resources for the second implementation.

See how one scenario includes more steps and processes than the other? I'm not saying Mantle isn't easy to code for, far from it. How easy it is isn't relevant. It still requires more work to support two API's than one.

 

[OCGuy]

Because anything I've read has said the opposite - ie Mantle is *not* difficult to develop for. If the multi-threading was a basic copy/paste job from DX12 then that would be half the job done already.

Any idea why it has missed the first two game launches by so much?

 

[SiliconWars]

Mutant - stands to reason that if devs are currently willing to code specifically for Mantle *now*, then more devs would be more willing to when half the job (CPU multi-threading) is already done in DX12.

OCGuy - because Mantle is still in beta and I've never yet seen any bug-free software released since "Hello World!".

 

[OCGuy]

Mantle is in beta? Or they are using Beta drivers? The slides I read talked about how simple it was.

Why would a API need a beta if it is so easy to port from consoles? Why would a developer not on the Gaming Evolved payroll want to pay man hours to beta test?

The facts will remain the same until non-sponsored games start popping up.

 

[SiliconWars]

Mantle is in beta? Or they are using Beta drivers? The slides I read talked about how simple it was.

Why would a API need a beta if it is so easy to port from consoles? Why would a developer not on the Gaming Evolved payroll want to pay man hours to beta test?

The facts will remain the same until non-sponsored games start popping up.

If DX multi-threading was so simple then why did they get it so wrong before? Where is it now? Microsoft doesn't even have anything remotely close to a DX12 beta - let alone any games ready for developing on it.

 

[OCGuy]

DX 12? DX11.2 is the big leap. DX works fine for me. There is a slim chance I would ever try Mantle for BF4 again, it looked terrible and stuttered.

 

[VulgarDisplay]

Software development must be pretty easy since thousands of people do it. There is no reason for them to beta test according to your posts logic. Hop off the amd hate train and at least be rational when you post.

I was actually under the impression that mantle is alpha software with beta drivers. I thought at GPU 14 they said mantle us alpha currently and sometime late 2014 or 2015 it will go beta and be opened up for other developers who want to try it.