The AMD Mantle Thread

[cmdrdredd]

Problem with your theory is the developers asked AMD to do this. Why would they want AMD to release a new API with lower-level access (something they begged for for years) and then not use it? It's not logical. Secondly, the game development market now consists of 2 leading consoles which are all powered by GCN and AMD will use GCN for at least another 3-4 years I bet. You know how large that userbase will become soon? Don't forget the chicken and egg scenario. If you don't release a lower-level API targeting a specific GPU architecture, developers will never use it. Therefore, you have to take a risk. AMD now has PS4/XB1 locked in, HD7000 series and Rx 200 series. That's a whole lot of GCN parts and then 20nm, 16nm, etc.

You think MS and Sony wouldn't want "free" 20-30% performance increase for their consoles from the very company which is providing them with the APUs?

And new consoles coming out, well I don't think we have to worry about that for another 7 years. We don't even know if next gen consoles will be physical or subscription-based model from the cloud. What about from now until PS5/XB2? That's a long time. Mantle is here and now, not in 7 years.

You aren't going to get any free performance from consoles. Not this absurd 30% number especially. Sony and Microsoft already have an api that does this. They have talked about it for a while, developers have talked about it for a while, and there were games almost done before mantle.

Why can't anyone see that this is dice and ea more than amd? If it was any different they would have note than one single developer with anything to show or say.

 

[cmdrdredd]

I find it fun how people seem to find so incredible that AMD can rip off whatever low APIs Xbone and PS4 are using.

And if the DX and OpenGL ES APIs in the Xbone and PS4 were the same as the computer DX and OpenGL ES then everyone could port the optimizations for the desktop they use in the consoles and the we could run games like BF3 in a Geforce 7800GTX with an Athlon X2.

So the logic dictates that there is something different in the consoles and that is the low level API.

And if that low level API can interact with the GCN GPU in the consoles it can also interact with the GCN in the PCs via the mantle API.

The api in consoles is built by Sony and Microsoft, not amd.

 

[ShintaiDK]

Problem with your theory is the developers asked AMD to do this.

By developers you mean DICE.

 

[GaiaHunter]

The api in consoles is built by Sony and Microsoft, not amd.

Sure, MS and Sony just guess how the API will communicate with the hardware without AMD being involved at all...

 

[Kenmitch]

Sure, MS and Sony just guess how the API will communicate with the hardware without AMD being involved at all...

Probably ask both of their API'S who's your daddy and they'll say Mantle.

 

[BrightCandle]

Ever since Microsoft added the context queue with DX10 to separate the two parts of the GPU drivers we have had latency issues associated with DirectX. The idea of queuing up the draw commands into a 3 long buffer before the card and then again afterwards having vsync buffers as well is a lot of designed in latency. The C++ interface is quite limited in multithreaded support and since that tends to dominate the usage in most games it does seem reasonable to want to remove the limitation there. The fixed pipeline representation in DX is also overly strict compared to what the cards can now do and since most games are now bypassing the fixed function lighting it seems reasonable to want to expand the API.

I think this release says a lot about the way DX has gone and what NVidia and AMD and the big developers wanted but didn't get negotiating with Microsoft. I wouldn't worry too much about the draw call performance increase as its not something they can use until DX/openGL are gone for gaming purposes anyway but it might give performance boosts at some point in the future. But the other changes like better threading support and skipping parts of the fixed function might very well bring a true change and performance benefits on todays CPUs and GPUs.

But the nagging feeling I have is that a low level API is bad overall for PC gaming. Its not just that its AMD only, its GCN with the current generation cards only. If we add a new fixed function or expand what Shaders can do this API could become obsolete or gain its own set of cruft and poor performing aspects. The abstract layer of DX and openGL allows the GPUs to keep getting better and changing their architecture which has served us well for well over a decade. A low level API might force them into keeping backwards compatibility in a way that is detrimental to the industries hardware development.

I need to see what this API looks like to know how much of a problem its going to be. Its either going to allow NVidia and Intel to implement it as well and hence its just an openGL/DX competitor updated for modern concerns or its really a low level API and AMD will have trouble supporting it into the future as well as their competitors. I am hoping its more a competitor that has more flexibility that the big developers wanted than a true LL API.

 

[VulgarDisplay]

The api in consoles is built by Sony and Microsoft, not amd.

From the Anandtech article that apparently no one bothered to read:

What’s not being said, but what becomes increasingly hinted at as we read through AMD’s material, is not just that Mantle is a low level API, but rather Mantle is the low level API. As in it’s either a direct copy or a very close derivative of the Xbox One’s low level graphics API. All of the pieces are there; AMD will tell you from the start that Mantle is designed to leverage the optimization work done for games on the next generation consoles, and furthermore Mantle can even use the Direct3D High Level Shader Language (HLSL), the high level shader language Xbox One shaders will be coded against in the first place.

http://www.anandtech.com/show/7371/understanding-amds-mantle-a-lowlevel-graphics-api-for-gcn

The article is also written kind of strange, and by that I mean I think when they say Xbox One they mean both next gen consoles. Either way, Ryan Smith who is far more knowledgeable about GPU's than most of the people posting on these forums seems to think it's "the API." On top of that AMD's twitter said that it's an "excellent summary" of what mantle is. They wouldn't say an excellent summary if it was wrong.

 

[Shivansps]

But what AMD said Mantle was to allow console style low level access on PC, not that ports going to be any better, not that a game that runs in a console its auto-optimised for pc, for that you need to use the exact same API as in console(Sony/MS privative, "similar" is not enoght) and even still its not the same doing it on a console than in a pc, you may end up with negative gains on certain AMD hardware too.

ports are horrible because they are cheap, you can do a good port if you wish, Mantle its not going to help with that, and i assure you that will end in even horrible DX/OpenGL performance.

 

[VulgarDisplay]

By developers you mean DICE.

Seriously? wow....

 

[SiliconWars]

http://translate.google.com/transla...-kaarten-mantle-prestaties-en-compatibiliteit

Skynner: "Let me say this about: we will not develop completely new API there 3 or 4 percent performance gain to be gained. The performance gain will be significant. "

Skynner: "That is the advantage of our Unified Gaming strategy. When it became known that we had let Nvidia know they were not important and these deals are not even wanted. The console business We believe it is just enormously important for gamers and it is a good thing when the console and PC world come close together. "

Bad translation but might be somthing like "when AMD won the contracts Nvidia said it wasn't important, and didn't really want them anyway".

 

[Saylick]

If this is true...

I wonder what kind of framerates they were getting.

EDIT: Oops, didn't notice this was a re-post.

Care to explain because I have no idea what you are getting at.

I meant that there wasn't going to be any specs unveiled at the pre-order date, so really, there isn't an NDA lifting at all.

 

[gorobei]

I think the real strategy will rest upon getting Mantle baked into various engines that are designed for cross platform use - Frostbite 3 will have it baked in starting in December for all games to use, and Frostbite 3 will be in nearly every high profile EA release in 2014, including Dragon Age: Inquisition. I think it's safe to state that Dragon Age 3 using Mantle is pretty much a certainty, since Frostbite 3 will have the ability baked in. If you remember, this year EA consolidated their dev studios heavily and commented that most of their developers would be using Frostbite 3 to power their games. So that would indicate multiple EA games using Frostbite 3 in 2014.

Aside from that, AMD is probably trying to get other engines to use it. The thing is, very few developers build their own "from scratch" engines so we have a situation where a few engines power the majority of the games. So far, despite the aging nature of Unreal Engine - UE3 does power quite a few games. Anyway, I'd imagine AMD would focus on implementing mantle in the 3-4 game engines that power 80%+ of games, and Crytek has mentioned on Twitter that they're looking into Mantle (this may make sense since AMD has developer relations with Crytek now). CryEngine 3 using Mantle would be a significant milestone, UE4 would be a big win too but so far Epic has not commented on Mantle. I should add that Activision is also receptive to Mantle, although let's not kid ourselves - Activision games have never pushed the "boundary" so to speak in terms of graphics.

If AMD can manage to get Mantle into 2-3 of the game engines that power 80% of games, well that would be the winning business strategy. They could presumably power a TON of games with little effort on behalf of any developer - and like I said, those 2-3 game engines power the vast majority of games. Developers are less willing to spend time creating an engine and generally license them these days. I'm 99% sure that AMD will go for this strategy.

While that probably is their strategy, it isnt an automatic choice to use mantle for most devs.

Remember epic just signed a deal/declared that UE4 would natively support a bunch of nv features including the apex physics stuff. So all the dev has to do is make a few library calls to be able to use some of the physx capability. That is a tempting freebee if your arent shooting for gameplay heavily enough defined by physics to write your own dynamics engine. It just means that the decision to use the nv-apex stuff falls to the game art director and whether or not that little bit of swirling mist is significant to his game esthetic.

Mantle in and of itself is a sort of free performance boost to your poly/framerate budget while truesound may let you offload your audio budget to the newer cards later on.

Only devs pushing bleeding edge graphics (dice, crytek, whoever does witcher, etc) will be looking at "turning the dial to 11/running at 150%/going to ludicrous speed"

Other devs may use mantle to expand the amount of lower end hardware that can play their game, thus more sales. So a game like torchlight2, or a f2p planetside2 might be playable on a jaguar or kaveri tablet.

But for the bulk of devs that just want overall compatibility and dont have the most demanding lighting there wont be a huge incentive to do the recode of the dx calls to mantle calls.

That may change when true radiosity and 4k and VR become the must haves.

 

[Shivansps]

What looks fishy to me its the timeline, September, Mantle hype, October new Radeons sales, November Mantle details.

I wonder why they cant give the details now, and why only DICE was named if lots of devs wanted it.

 

[sandorski]

What looks fishy to me its the timeline, September, Mantle hype, October new Radeons sales, November Mantle details.

I wonder why they cant give the details now, and why only DICE was named if lots of devs wanted it.

In a few months we'll know much more.

 

[RussianSensation]

By developers you mean DICE.

No. There is already an article at bit-tech posted on our forums and comments by developers over the years regarding DX API being inefficient -- multiple developers on multiple times have been asking for a lower level API to extract more performance. DICE just happens to be the first guinea pig to test it out.

But what AMD said Mantle was to allow console style low level access on PC, not that ports going to be any better, not that a game that runs in a console its auto-optimised for pc, for that you need to use the exact same API as in console(Sony/MS privative, "similar" is not enoght) and even still its not the same doing it on a console than in a pc, you may end up with negative gains on certain AMD hardware too.

ports are horrible because they are cheap, you can do a good port if you wish, Mantle its not going to help with that, and i assure you that will end in even horrible DX/OpenGL performance.

As VulgarDisplay noted, per AT's article on Mantle, the XB1's lower-level API is Mantle. You make the game with Mantle API on XB1 and then port it to PC with much of the same code intact. That's where you get the optimization for PC hardware, albeit for GCN hardware only. In theory it should make for a much more optimized console-to-PC port assuming Mantle is actually faster than DX port. Of course if games are made on the PC first and then ported to the console, then this wouldn't help much unless the PC developer makes the original game with Mantle API as well (such as the case with BF4). But if XB1's lower level API is mantle, coding a PC game in Mantle would make porting from PC to console easier too. In both cases even if you port console to PC or PC to console, you transfer the code since Mantle can be used for both.

 

[BoFox]

Ever since Microsoft added the context queue with DX10 to separate the two parts of the GPU drivers we have had latency issues associated with DirectX. The idea of queuing up the draw commands into a 3 long buffer before the card and then again afterwards having vsync buffers as well is a lot of designed in latency. The C++ interface is quite limited in multithreaded support and since that tends to dominate the usage in most games it does seem reasonable to want to remove the limitation there. The fixed pipeline representation in DX is also overly strict compared to what the cards can now do and since most games are now bypassing the fixed function lighting it seems reasonable to want to expand the API.

I think this release says a lot about the way DX has gone and what NVidia and AMD and the big developers wanted but didn't get negotiating with Microsoft. I wouldn't worry too much about the draw call performance increase as its not something they can use until DX/openGL are gone for gaming purposes anyway but it might give performance boosts at some point in the future. But the other changes like better threading support and skipping parts of the fixed function might very well bring a true change and performance benefits on todays CPUs and GPUs.

But the nagging feeling I have is that a low level API is bad overall for PC gaming. Its not just that its AMD only, its GCN with the current generation cards only. If we add a new fixed function or expand what Shaders can do this API could become obsolete or gain its own set of cruft and poor performing aspects. The abstract layer of DX and openGL allows the GPUs to keep getting better and changing their architecture which has served us well for well over a decade. A low level API might force them into keeping backwards compatibility in a way that is detrimental to the industries hardware development.

I need to see what this API looks like to know how much of a problem its going to be. Its either going to allow NVidia and Intel to implement it as well and hence its just an openGL/DX competitor updated for modern concerns or its really a low level API and AMD will have trouble supporting it into the future as well as their competitors. I am hoping its more a competitor that has more flexibility that the big developers wanted than a true LL API.

That is what I originally thought about NV's PhysX, that it was bad overall for PC gaming. ATI introduced tessellation with Unreal Tournament (TruForm), tried to promote it a bit later on (although they dropped hardware TruForm with 9700Pro) with a Ruby demo for HD 2900XT yet Nvidia basically dropped tessellation for nearly a decade before it became a standard with DX11. GPU-based physics was just a tad bit more successful, but still.. it's not quite yet ubiquitous.

However, Mantle is more like Nvidia's CUDA than a more specific function like PhysX (which uses NV's CUDA). We have had CUDA for nearly 7 years now, and an 8800GTX can pretty much run PhysX just as well as it ever did (within its own capacity). I know that Just Cause 2's CUDA filters require GT200 rather than G80 (IIRC), but the point is that with GCN being the basic requirement of Mantle, I think AMD will simply be building upon GCN without changing too many things for the upcoming console generation of games. It will probably only be GCN 2.0, 2.1 and so on forward, without moving onto a completely new arch - given that GCN is in its flexible essence forward-looking from a rather different VLIW-based arch.

It's not like as if the console API is only going to be good for 2-3 years of PC ports and then start to hold everything back. Look at Skyrim, which is like 5-10x as graphically stunning as Oblivion. Skyrim came out many years later on the aging consoles, yet showed us how the same console could still show a vastly better-looking game. The devs were still finding more efficient ways of coding to the metal after all these years, and getting amazing results. Bottom line is that we still got SKYRIM on these ancient consoles. GTA IV is another story - more to do with draw calls and memory management (that could have drastically benefited with an API of the same nature for PC's, encouraging the dev to port easily to). Look at how long it took for the devs to improve Skyrim performance with patches for the PC version.

As for comparing it to Glide - heck, some games ran SEVERAL times faster on Glide than on DirectX (UT '99), or even OpenGL. DX7 and below were abysmal - just look at how Half Life 1 ran compared to OpenGL on the older rigs. I don't expect Mantle to run THAT much faster than DX11, DX11.1 or DX11.2, but a minimum of 10% speed increase is a modest expectation, if I had to guess. The difference over DX11.2 might be lower than DX11.0 (which is obviously one of the reasons why M$ just had to do DX11.2 for XBone). :sneaky:

Imagine Kaveri APU (Steamroller-based CPU with GCN-based GPU) coming on socket FM2+ in Q1 2014, with more benefit from Mantle than all other Radeon cards, thanks to the huge efficiency boost with CPU and bandwidth utilization! There could indeed be more than 20% gain to be seen there, easily.

One thing for sure is that AMD is now working more closely with game devs than ever before. Only if ATI did not give up with Truform, but it did look funny in its infancy. AMD has been maturing as of late, learning from Nvidia and from their own failures. I just hope AMD keeps it on the up and up, without the 65nm Athlon (weaker than 90nm), Phenom 1, Bulldozer, HD 4890 using 800sp instead of 960) etc.. moments happening again ever - because AMD really needs all the constructive momentum they can possibly achieve.

 

[0___________0]

I'm very skeptical of Ryan's conclusion that Mantle is "the API" on the XB1, for a couple of reasons.

I think his primary assumption is flawed, he's assuming there's two different API's on the XB1. This isn't the case with the 360, you only get D3D. Distinction, this is not the same as the DX you get on your PC, the 360's is fine tuned, exposes basically all functionality on the GPU, and is lower level. The consoles also don't have expensive transitions from user->kernel mode, an expensive HAL, etc; there's more to the consoles' advantage then the graphics API. I don't see why Microsoft all of a sudden decides that their own API isn't good enough, they can make D3D as low level as they want for the XB1. Now, all of that said, the D3D API on the 360 wasn't as lean as it could be, that was intentional though; but this time around MS doesn't need to impose those restrictions. They can always provide a wrapper to allow easier development while preserving a lower level API for performance efficiency.

My other issue with it is how is it even on the XB1? People are tossing around the Carmack quote about MS potentially being hostile to Mantle, so why would they allow it on the XB1? All of the API's they employ on the consoles have got their own driver this time around and I'd expect the same for AMD's implementation of Mantle on the consoles. So why would MS put that driver on their console? Perhaps AMD just had that much leverage when they negotiated the contract.

Highly unlikely in my opinion that an API developed exclusively by AMD and DICE is on the XB1, there's no indication MS was involved, at all.

I'm also curious why he thinks it's only on the XB1 and not the PS4; I'm going to assume that's because if I recall correctly there's already two API's of the PS4 of varying closeness with the hardware; or at least a wrapper. If AMD could get it on the XB1, I don't see any reason why they couldn't get Sony to put it on the PS4 either. There's also no necessity for the XB1 to have another API besides DX.

If it is already on the XB1, I find it a great curiosity that AMD has neglected to mention it, that would be great support, two of the 3 primary development platforms sharing it is significant. It also seems like only the PC version of BF4 is getting Mantle. That's very strange if Mantle is already present as "the API" on the XB1.

I'm not going to bother speculating on exact performance, it's obviously going to be faster than what we're getting with DX right now, assuming it isn't horribly bugged; nothing besides that is for certain.

EDIT: I'm not convinced by a PR drone's tweet either. I seriously doubt they did a detailed vetting of that article, since they probably have little, or more likely, no understanding of more technical matters. Besides, why just a quasi confirmation like that? Makes no sense.

 

[imaheadcase]

I would like to point out that BF3 and BF4 both use Nvidia API they made as well. Which is nice and all for both camps using those cards. Pointing that out because some think its just AMD is going to benefit in those titles.

 

[Final8ty]

I would like to point out that BF3 and BF4 both use Nvidia API they made as well. Which is nice and all for both camps using those cards. Pointing that out because some think its just AMD is going to benefit in those titles.

Indeed.

AMD’s Mantle API is currently being integrated in Frostbite 3 based Battlefield 4 which is without a doubt the most biggest title coming out this year after GTA V. Such is its fame that AMD even bundled their latest and top new Radeon R9 290X graphic card with the new title and those of you lucky enough to pre-order the GPU now would be able to redeem the game at no additional cost. So back to the API talk, currently developers have to operate through DirectX and OpenGL APIs to make games work but this doesn’t fully unleash the hardware capabilities of a PC nor do they allow ease of development to developers.

The AMD Mantle API is being exclusively developed for GCN enabled Radeon graphic cards. This would allow developers to dig deep into the metal to bring console-level optimizations through ease of programming and faster optimizations over a coherent GCN chip architecture. This means that we would see better performance on the entire GCN architecture enabled AMD graphic card lineup ranging from the top Radeon R9 290X to the bottom R7 250X.

But Frostbite 3 is more than that as John Anderson, the lead guy behind the team at DICE in developing AMD’s Mantle API said on his twitter profile that Battlefield 4 would also feature NVIDIA’s NVAPI support as it did in Battlefield 3. While the optimizations many not be as great with NVIDIA’s API as with AMD’s Mantle, its still worth noting that atleast DICE is supporting both the Red and Green team graphic cards which means PC optimizations at both ends. In addition to this, you will have the option to select between using Mantle and DirectX 11 if you are using a GCN enabled GPU. Frostbite 3 is on the road to become one of the new mammoth tech engines in the gaming industry powering a portfolio of 15 AAA titles which include the upcoming:

Battlefield 4
Command and Conquer
Mirrors Edge
Plants Vs Zombies: Garden Warfare
Need For Speed: Rivals
Dragon Age: Inquisition
Star Wars: Battlefront
Mass Effect (New Title in the Franchise

http://wccftech.com/battlefield-4-frostbite-3-support-amd-mantle-nvidia-nvapi-apis-pc-optimizations/

 

[imaheadcase]

Speaking of which, couldn't nvidia simply update its api to take instant advantage to a game like bf4? Or does the game engine need the specifics programmed into it to use the newer one?

I could see nvidia simply updating its api to compete with mantle. Surely what AMD knows, nvidia knows.

Or does that require new video cards/software update?

 

[gorobei]

I would like to point out that BF3 and BF4 both use Nvidia API they made as well. Which is nice and all for both camps using those cards. Pointing that out because some think its just AMD is going to benefit in those titles.

can you specify how or what features they are using?

all i could find on the nv page was this

NVAPI Feature Categories

Driver Management
Initialization and driver version controls.
GPU Management
Enumeration of physical and logical GPUs. Thermal and Cooling controls.
Display Management
Enumeration of NVDIA displays, display postion and timings controls.
GPU Topology
Ability to enable SLI and Hybrid GPU topologies.*
Frame Rendering
Ability to control Video and DX rendering not available in DX runtime.*
System Management
Ability to query chipset and system specific information.
HDTV Controls
HDTV format and overscan controls.*
Video Controls
Extended video engine controls.*
Connecting and Configuring Monitors
Ability to set views on multiple target monitors.
GPU Overclocking
GPU overclocking APIs allows apps to run apps at maximum possible clocks. *

none of this looks like any kind of low level access to the architecture.

 

[Grooveriding]

can you specify how or what features they are using?

all i could find on the nv page was this


none of this looks like any kind of low level access to the architecture.

I would guess the difference is they addresss some features of nvidia GPUs whereas with AMD's Mantle it is going to be an entire render path for AMD users, where you'll be selecting DX11.1 or Mantle as the renderer in the game's options. While nvidia will continue to use DX11 and whatever architecture specific optimizations DICE makes for nvidia cards in the game.

It sort of sounds like the DICE employee is playing PR fancy with words to avoid highlighting that the AMD specific Mantle render path in BF4 is going a much more robust and specific nod to one vendor.

 

[Cloudfire777]

Indeed.

http://wccftech.com/battlefield-4-frostbite-3-support-amd-mantle-nvidia-nvapi-apis-pc-optimizations/

Its quite funny.

BF3 is a gaming evolved title and was said to play soo much better on AMD cards. But that didnt turn out to be true.

Since DICE now confirm that they used NVAPI on BF3, you could think that it already plays a role in todays games. Nvidia have much more cash than AMD not to mention way bigger software team, and since they have their own API, I dont think Nvidia have anything to fear in the future.

IF Mantle becomes a success.



Good to hear that DICE also will be using NVAPI on BF4

 

[blackened23]

NVAPI was also used in bf3, it is not a low level API.

Driver Management
Initialization and driver version controls.
GPU Management
Enumeration of physical and logical GPUs. Thermal and Cooling controls.
Display Management
Enumeration of NVDIA displays, display postion and timings controls.
GPU Topology
Ability to enable SLI and Hybrid GPU topologies.*
Frame Rendering
Ability to control Video and DX rendering not available in DX runtime.*
System Management
Ability to query chipset and system specific information.
HDTV Controls
HDTV format and overscan controls.*
Video Controls
Extended video engine controls.*
Connecting and Configuring Monitors
Ability to set views on multiple target monitors.
GPU Overclocking
GPU overclocking APIs allows apps to run apps at maximum possible clocks. *

None of these are low level "to the metal" functions so NVAPI is different than Mantle in this respect. And this isn't surprising since BF3 also used NVAPI.

 

[Final8ty]

Its quite funny.

BF3 is a gaming evolved title and was said to play soo much better on AMD cards. But that didnt turn out to be true.

Since DICE now confirm that they used NVAPI on BF3, you could think that it already plays a role in todays games. Nvidia have much more cash than AMD not to mention way bigger software team, and since they have their own API, I dont think Nvidia have anything to fear in the future.

IF Mantle becomes a success.



Good to hear that DICE also will be using NVAPI on BF4

It was well known the DICE worked with both NV and AMD and both were promoting BF3 so no surprise at all and im sure SLI was being promoted by NV with BF3 saying this is what you need for ULTRA.