So what's your take on Mantle

Mantle is a bad thing for the industry, its proprietary by a minority brand. The sad thing is it actually has a chance of taking off thanks to the console monopoly.

The way I see there is only 1 good thing that can come from mantle. It could motivate Microsoft and the openGL collective to work on issues they have with those APIs. But other than that all it is going to do is divide developer time and create more problems with compatibility but long term it probably will not solidify itself a s permanent player.
 
Mantle is a bad thing for the industry, its proprietary by a minority brand. The sad thing is it actually has a chance of taking off thanks to the console monopoly.

The way I see there is only 1 good thing that can come from mantle. It could motivate Microsoft and the openGL collective to work on issues they have with those APIs. But other than that all it is going to do is divide developer time and create more problems with compatibility but long term it probably will not solidify itself a s permanent player.
Can one rightly call AMD a minority brand if they exist in all the millions of upcoming consoles which are more PC-like than ever? With Mantle that distinction is all but wiped out so far as software development is concerned.

The only bad thing that can come of Mantle's success would be relegating Nvidia to minority status, and that would only be bad in the eyes of Nvidia themselves and fanboys. Someone has to be in the minority, right? I don't see why it would be such a tragedy if AMD took the lead. There is little chance they would outright put Nvidia out of business.
 
only thing mantle will do is give amd a nice edge up on the competition ,in games that uses it.
and not like the industry is giving up on dx. Might make some ppl green tho :p

i wonder how long it will supports the 5000 6000 series .... last big release ,4000 and down went the way of the dodo. then there would only be GNC.... *insert there can only be one quote*
 
Can one rightly call AMD a minority brand if they exist in all the millions of upcoming consoles which are more PC-like than ever? With Mantle that distinction is all but wiped out so far as software development is concerned.

The only bad thing that can come of Mantle's success would be relegating Nvidia to minority status, and that would only be bad in the eyes of Nvidia themselves and fanboys. Someone has to be in the minority, right? I don't see why it would be such a tragedy if AMD took the lead. There is little chance they would outright put Nvidia out of business.

Yes they can because those are not PCs, in fact they only speak more to the problem as consoles have been the major problem with PC gaming, it has divided developer time, caused people to need to take sides etc.... People tend to forget all this really is a big giant shitty format war and in the end the customers are losing. And it used to be at least the format war was mostly console vs windows but now windows is under large attack. The only thing this will do is be one more negative people can point to and say, see look whats wrong with PC gaming if you don't have an AMD card you cant play this game or that game.

If AMD, NVidia and Intel could have worked together on this then it might have been worth something.
 
nvidia said they spend 1.5Billion a year on R&D i think they can spend a few hundred mill instead on the devs to get their own API widely accepted to compete if it will mean that much for a to the metal API that will give current and future gpu's a kick in the pants, or they can maybe put some of that money to good use and actually come up with a ground breaking new approach that blows amd and its mantle out of the atmosphere with pure brute force.
 
only thing mantle will do is give amd a nice edge up on the competition ,in games that uses it.
and not like the industry is giving up on dx. Might make some ppl green tho :p

i wonder how long it will supports the 5000 6000 series .... last big release ,4000 and down went the way of the dodo. then there would only be GNC.... *insert there can only be one quote*

Mantle does not support the 5000 or 6000 series, nor does it support any of AMD's current APUs.
 
If AMD, NVidia and Intel could have worked together on this then it might have been worth something.

If all of them got together to write an API it would be another OpenGL/DirectX clone. The point of low level API is specific to architecture, none of these players create similar architectures for this type of thing.

Right now we are to believe that developers were craving something like for both PC/Console type of development. I'll believe them until developers start coming out and begging for top level API's again.

This is a waiting game.

only thing mantle will do is give amd a nice edge up on the competition ,in games that uses it.
and not like the industry is giving up on dx. Might make some ppl green tho :p

i wonder how long it will supports the 5000 6000 series .... last big release ,4000 and down went the way of the dodo. then there would only be GNC.... *insert there can only be one quote*

You'd have to create different API's for each of those as they have different architectures. GCN was created with the intention of being used for a long long time. I think we finally understand why AMD wanted it that way.

nvidia said they spend 1.5Billion a year on R&D i think they can spend a few hundred mill instead on the devs to get their own API widely accepted to compete if it will mean that much for a to the metal API that will give current and future gpu's a kick in the pants, or they can maybe put some of that money to good use and actually come up with a ground breaking new approach that blows amd and its mantle out of the atmosphere with pure brute force.

I honestly believe the developers wanted Mantle for easier PC porting. I think if Nvidia creates a low level API such as Mantle, it would get used sparingly or when their pockets get deeper. Mantle will for sure put Nvidia is a much tighter spot then they were before.
 
nvidia said they spend 1.5Billion a year on R&D i think they can spend a few hundred mill instead on the devs to get their own API widely accepted to compete if it will mean that much for a to the metal API that will give current and future gpu's a kick in the pants, or they can maybe put some of that money to good use and actually come up with a ground breaking new approach that blows amd and its mantle out of the atmosphere with pure brute force.

Difference being AMD has gotten, and will continue to get paid for their efforts with Mantle. It's paid for by the consoles and is simply available to them for their own GPUs for only the cost of implementing it. It's already developed. It's also already being used in games for consoles. nVidia will have to develop a low level API and then give people a reason to use it, like they have to do with GPU PhysX. We would likely not see it used in any more games than we see GPU PhysX used in now.
 
Difference being AMD has gotten, and will continue to get paid for their efforts with Mantle. It's paid for by the consoles and is simply available to them for their own GPUs for only the cost of implementing it. It's already developed. It's also already being used in games for consoles. nVidia will have to develop a low level API and then give people a reason to use it, like they have to do with GPU PhysX. We would likely not see it used in any more games than we see GPU PhysX used in now.

Any proof of that? I would imagine that the consoles would have their own api designed to leverage all of their added features without any added bloat.
 
Any proof of that? I would imagine that the consoles would have their own api designed to leverage all of their added features without any added bloat.

If you watch 2:39 and out, you'll hear AMD say (I'm not native English, so its something like this):

"What is Mantle? Mantle is a new low-level programming interface with a very thin driver model, and it is also compatible with the existing shading languages like Microsoft HLSL, to make the job of porting existing games and assets really really easy".
http://www.youtube.com/watch?v=Xo1jJJGhneg

Dice also explains that Mantle will be cross-platform in their slides:

1380153080A5gFxFf9Dz_1_4.jpg

http://www.hardocp.com/news/2013/09/25/amd_introduces_mantle_at_gpu_14

Anandtech goes more into depth talking about the API similarities between consoles and porting it to PC:
http://anandtech.com/show/7371/understanding-amds-mantle-a-lowlevel-graphics-api-for-gcn
 
Yeah, I get that its cross-platform but are there any advantages as a developer over using existing apis.

There is a software developer that posts over at Rage3d pretty regularly. Here is a quick post that he made. Its something to think about.

It's unlikely as I see it; both Sony and MS made significant changes to the APU package so chances are Mantle won't suit the consoles where there is already an existing 'thin' API in place.

Link

Like most of the people commenting in this thread I'm no software developer. I don't think that the answer is as clear cut as many people want it to be.
 
Yeah, I get that its cross-platform but are there any advantages as a developer over using existing apis.

There is a software developer that posts over at Rage3d pretty regularly. Here is a quick post that he made. Its something to think about.



Link

Like most of the people commenting in this thread I'm no software developer. I don't think that the answer is as clear cut as many people want it to be.

What made Mantle possible, is the GCN hardware that exist in both the console, made by AMD, and Mantle only works on AMD GCN hardware. That this developer thinks AMD is releasing Mantle, not knowing the low-level close to AMD metal API in both consoles, is a bit of a stretch, don't you think?

Do you really think that AMD spent 2 years developing Mantle and Consoles, and don't stay updated on the progress? :)
 
What made Mantle possible, is the GCN hardware that exist in both the console, made by AMD, and Mantle only works on AMD GCN hardware. That this developer thinks AMD is releasing Mantle, not knowing the low-level close to AMD metal API in both consoles, is a bit of a stretch, don't you think?

Do you really think that AMD spent 2 years developing Mantle and Consoles, and don't stay updated on the progress? :)

Again, you're making assumptions without any experience in software development.

Do you really think that the custom apis used in consoles aren't already low level? Would mantle not need excess bloat in order to support all of these different platforms compared to existing custom apis? There are other aspects that these consoles have cameras, screens in the controller, etc that I'm sure are already handled by an thin existing api.

Again, I'm sure that there is more to it than "cross-platform" so thats what everyone will use. I don't think that the answer is that easy. Again, in practice I'm sure that things change. We can all try and be arm chair quarterbacks but without any experience in the field its hard to make any real opinions.

I'm not saying that Mantle won't be used on the pc. I am saying that it might not be the most used api on the consoles.
 
Here is another very good post that I saw in that Rage3d thread.

As bobvodka indicated both the PS4 and Xbone have their own API which I would think take advantage of the whole architecture such as camera, networking, game controller sound etc. Mantle appears to be GCN orientated for now and was co-developed by a few developers with AMD for their needs.

I would also think PS4 and Xbone APIs would be closer to the metal since it is a console with relatively fixed specs for the hardware capability for a significant length of time.

Will developers use Mantle and other console APIs together? Are the developer tools as good or better with Mantle then with what Sony and Microsoft has come up with?

If a console developer used Mantle how could they get help from Sony or Microsoft? AMD?

Will a developer use Mantle for %30 of the audience? I don't know, but AMD GCN arch has been around awhile, APU's of the future will use GCN and the numbers will grow rather fast. Plus getting those APUs working well for gaming, Mantle seems very interesting and a forward looking solution.

Making a game that shines, even on modest hardware I would think would be very enticing for some developers.

Nvidia making their own derivative of Mantle would probably make the API stick where the two together would represent the serious PC gaming industry. BF4 could be a game changer if performance and effects blows everyone else's away. I would like to see Nvidia beat AMD on something using Mantle - that would be fun and give Nvidia a hell a lot of credit in their capability.

For most it will be raw horsepower for today games using current standards.

All I was pointing out is that, unlike assumptions being made all over the place, Mantle isn't going to be a 'drop in' API; it'll still be another code path which developers will have to maintain along side the existing ones.

That's all.

Its something to think about.
 
I'm really looking to see what Mantle can offer on low to mid level gcn hardware. I think that' where we'll see the most benefits. If I can play my games on a $150 card with uber eye candy vs.a $300-$500 card more power to them...
 
Again, you're making assumptions without any experience in software development.

Do you really think that the custom apis used in consoles aren't already low level? Would mantle not need excess bloat in order to support all of these different platforms compared to existing custom apis? There are other aspects that these consoles have cameras, screens in the controller, etc that I'm sure are already handled by an thin existing api.

Again, I'm sure that there is more to it than "cross-platform" so thats what everyone will use. I don't think that the answer is that easy. Again, in practice I'm sure that things change. We can all try and be arm chair quarterbacks but without any experience in the field its hard to make any real opinions.

I'm not saying that Mantle won't be used on the pc. I am saying that it might not be the most used api on the consoles.

No, I'm just saying that this developers assumptions are a bit of a stretch. He assumes that MS and Sony made changes that AMD are not aware of and haven't been aware of from the 2 years of developement up to the summit where they now introduced Mantle.

Thats a huge stretch by him, dont you think?
 
On a side note, I'm curious to know which shader language will the PS4 use. PS3 uses nvidia's CG and the xbox has HLSL. I'll have to assume GLSL

Back to topic, I wonder if MS would endorse Mantel for Xbone as its pushing DX11. Since the PS4 is supossedly faster, this would only widen the gap.
 
No, I'm just saying that this developers assumptions are a bit of a stretch. He assumes that MS and Sony made changes that AMD are not aware of and haven't been aware of from the 2 years of developement up to the summit where they now introduced Mantle.

Thats a huge stretch by him, dont you think?

Thats not at all what he is saying. Click the link and read a few posts.

On a side note, I'm curious to know which shader language will the PS4 use. PS3 uses nvidia's CG and the xbox has HLSL. I'll have to assume GLSL

Back to topic, I wonder if MS would endorse Mantel for Xbone as its pushing DX11. Since the PS4 is supossedly faster, this would only widen the gap.


I'm sure that Microsoft will have a custom api tailored for the console.
 
Disclaimer: I've done quite a bit of GPGPU programming and have dabbled in OpenGL a tiny bit. Like the rest of you, this all just speculation and is just what makes the most sense to me from my point of view.

Again, you're making assumptions without any experience in software development.

Do you really think that the custom apis used in consoles aren't already low level? Would mantle not need excess bloat in order to support all of these different platforms compared to existing custom apis?

Mantle wouldn't need any excess bloat to support these different platforms, simply because Mantle isn't meant to support them. As far as I understand, Mantle is only meant to be used in the rendering path, whether the code is meant to be executed on either side of the APU(cpu or gpu). I don't think that existing APIs have the capability to take full advantage of HSA or are programmed to utilize the unified address space, which cuts a ton of fat out of the previous programming model. Mantle is just AMDs solution for the rendering path.

There are other aspects that these consoles have cameras, screens in the controller, etc that I'm sure are already handled by an thin existing api.

I don't see any way that a console can get by without having it's own custom API, just to handle things like this. And obviously, console aren't restricted to only having one API, that's just not how software works. Each platform's custom API will make calls to the Mantle API, instead of it's own low-level rendering solution.

Again, I'm sure that there is more to it than "cross-platform" so thats what everyone will use. I don't think that the answer is that easy. Again, in practice I'm sure that things change. We can all try and be arm chair quarterbacks but without any experience in the field its hard to make any real opinions.

I'm not saying that Mantle won't be used on the pc. I am saying that it might not be the most used api on the consoles.

Since they've been working in conjunction with DICE on Mantle, I'd imagine they have had a lot of input from everyone in the pipeline, including the hardware and driver guys at AMD as well as from the end users at DICE. I'd be really surprised if another API developed independently could even come close to utilizing the console hardware as efficiently.

The beauty of this is that AMD has placed hardware in both the consoles and PCs. From Mantle's point of view, everything should look the same, it probably just sees a compute device with some # of GCN cores, etc... The advantage is two-fold then; because of Mantle, games will look fantastic on AMD PC hardware, plus, the studios have minimal effort to jump from the console into the PC market, which should translate into a little more profit.
 
Thats not at all what he is saying. Click the link and read a few posts.

It sure sounds like he's saying that. Unless he is unaware of AMD's presentation of Mantle as an API to port existing games and assets, and instead believe that AMD created Mantle to replace PS4 and XBone's API's.

Thats the thing I and many others are excited about with Mantle. Games made for consoles are made for the fixed hardware. In AMDs presentation, they speak about how you get less with more hardware when porting console games over to PC and how Mantle will give you more with more hardware. "Have your hardware cake and eat it too".

They only spoke of the XBone support, where they say Mantle support the low-level optimizations and the XBone's HLSL shader language. There is very little details about Mantle now and we'll probably not get the whole picture before November.

But, having my hardware cake and eat it too? Hell yeah! Games are already using low-level optimizations on PC. As you can see in the other thread, its being done for Nvidia too via NVAPI. Do I want more of that? Hell yeah! Do I want more of that for Nvidia too? Hell Yeah! Do I give a crap if developers have to spend effort optimizing the games so I get more performance out of my hardware? Hell no!
 
Since they've been working in conjunction with DICE on Mantle, I'd imagine they have had a lot of input from everyone in the pipeline, including the hardware and driver guys at AMD as well as from the end users at DICE. I'd be really surprised if another API developed independently could even come close to utilizing the console hardware as efficiently.

You're surprised that an api designed specifically for the console by the people who designed the console would work as efficiently?
 
NVIDIA is finished if they try to make their own API. Developers have no choice but to use Mantle, because AMD has control of the consoles, but they have absolutely no reason to bother with Nvidia's API...

Actually, MS and Sony have control of their consoles. They simply use AMD hardware.

Mantle is too early to say how well it is and how well supported it will be. Glide was suppose to be the next great thing too, and was hardly use and ended up being abandoned.
 
If all of them got together to write an API it would be another OpenGL/DirectX clone. The point of low level API is specific to architecture, none of these players create similar architectures for this type of thing.

Right now we are to believe that developers were craving something like for both PC/Console type of development. I'll believe them until developers start coming out and begging for top level API's again.

This is a waiting game.



You'd have to create different API's for each of those as they have different architectures. GCN was created with the intention of being used for a long long time. I think we finally understand why AMD wanted it that way.



I honestly believe the developers wanted Mantle for easier PC porting. I think if Nvidia creates a low level API such as Mantle, it would get used sparingly or when their pockets get deeper. Mantle will for sure put Nvidia is a much tighter spot then they were before.

They can get together and agree on a certain architecture or unit that will be consistent over their GPUs, then a low level API can be written that will work with all of their GPUs. IE make their metal more consistent between brands.
 
Mantle is too early to say how well it is and how well supported it will be. Glide was suppose to be the next great thing too, and was hardly use and ended up being abandoned.

That is just your uninformed opinion or like Harlan Ellison put it, "A fart in a wind tunnel".

The 2 major industry publishers WILL support Mantle.

EA
he9xbG2.jpg


Activision
ghthM0i.jpg
 
I find activision's comment quite distasteful... why do anything for you to increase your already substantial profits? AMD is already working with developers... but it does show everyone how corporate gaming studio execs think, especially when stuck with their big competitor having a leg up on them.

For Charlie, I would be interested to know if he didn't know either, or with only a select few in the know, held off writing anything, but from his free tidbit, he sees the huge potential!
 
More insight from an interview with Vr-zone: http://vr-zone.com/articles/mantling-alliances-ritchie-corpus-amd-interview/58215.html

The plan is, long term, once we’ve developed Mantle into a state where it’s stable and in a state where it can be shared openly [we’ll make it available]. The long term plan is to share and create the spec and SDK and make it widely available. Our thinking is: there’s nothing that says that someone else could develop their own version of Mantle and mirror what we’ve done in how to access the lower levels of their own silicon. I think what it does is it forges the way, the easiest way.

If you think about it, Mantle is truly a collaboration and I’ll tell you [DICE’s Johan Andersson] was at the forefront of that at the very beginning. A lot of the feedback on the development of Mantle came from him. We also solicited feedback from a lot of other partners that we haven’t announced yet. At this stage, Battlefield 4 and FrostBite 3 are the closest to deliver something today. I think, as I mentioned before, the goal would be to provide the spec and SDK publicly.

They also debunk the Steambox = Nvidia nonsense.
VR-Zone: Valve has recently announced that the Steam Box is in beta. In Valve’s announcing press release, it said how Nvidia is a big stakeholder in the Steam Box. What is AMD doing to counter or compete against this?

There is no counter.

The reason is we are working just as closely with Valve. I think the difference is one side of that conversation is being more vocal than the other. If you go to Valve right now and ask them which hardware partners are going to be the partner of choice, they won’t pick one or the other. They are going to be agnostic. They mention that it’s an open ecosystem.
 
Just saw this post on Rage3d. I figured that I would post it here for the people thinking that Mantle had anything to do with consoles.

Mantle is a PC API that allows developers to write direct to the hardware, similar to how is available on consoles, in HLSL. While the API's are different on each OS, using HLSL with direct h/w access on PC means removing bottlenecks and exposing hardware features more completely (partially resident textures, for example).
 

Of course Mantle have something to do with consoles. I think you misunderstand Jims post completely. Jim is explaining what Mantle is, he's not saying Mantle doesn't have anything to do with consoles. I don't think Jim would like it, if you spread around that he claimed Mantle has nothing to do with consoles.

There has been speculations (Ryan from Anandtech in Anandtechs article about Manlte) that "Mantle is the Xbox One’s low level API brought to the PC". That is however speculations, though Ryan has an educated guess in this.

What we DO know, is that Mantle is made to leverage optimization work from next-gen game consoles to PC (read last line in slide below). We also know that its compatible with HLSL for simplified porting. What we also know from Dice's presentation, is that Mantle will also be an API natively supported for windows in the Frostbite 3 Engine:

MantleBenefits_575px.jpg

MantleHLSL_575px.jpg

1380153080A5gFxFf9Dz_1_5.jpg


Oh yeah, we do know that Mantle have a lot to do about the new consoles. Read AMD's own slides. We also know that AMD created it to get a console-like API for PC. Mantle is made for PC users, if they use it on the consoles or not, we don't know. That would only be speculations at this point as shown with Anandtechs article.
 
There has been speculations (Ryan from Anandtech in Anandtechs article about Manlte) that "Mantle is the Xbox One’s low level API brought to the PC". That is however speculations, though Ryan has an educated guess in this.

Thats been debunked by many knowledgeable posters on B3D forums. Unless you have some hard evidence to back that up.

Everyone keeps repeating this and yet no one has said any such thing; a few sites have jumped to this conclusion and others have kept repeating it from there.

As I stated before;
- XBox uses a modified D3D11 path
- PS4 uses Gnm provided by, tweaked by and debugged by Sony

Both are tweaked and customised based on the hardware in the consoles which are NOT normal GCN parts as they have BOTH made changes to the APUs from stock parts at both CPU and GPU level.

Hell, AMD themselves keep saying 'PC API'...

(Frankly it's this kind of 'make **** up and repeat it' stuff which makes me laugh at tech 'journalist' and gain general amusemnet from the gaming population as a whole as NO ONE has bothered to fact check this stuff and yet people seem to think it is fact...)

Here is another great post on the topic by developer who works for a company that develops multi-platform titles. Yeah, I would think that a software developer working for a company that is developing games for these upcoming consoles might know a thing or two on the topic.

Yes, but sayign "Leveraging optimisation work" isn't the same as saying 'mantle on the console' - it just means that by opening up the hardware to the same level of access means that the same tricks can be applied to it.

So, once again, no one is saying 'mantle on the consoles' apart from people simply making **** up/assumptions who know nothing about the consoles and do not work with them. The APIs, as they exist on the consoles, have specific changes and optimisations for the custom work both companies have done - a generic GCN targetting solution would lose that ability.

Also, we have been working with those APIs for some time, no one is going to junk months of work to switch to an API which gives them no advantages on the hardware and maybe even removes features (even if MS and/or Sony allowed the lib on their OS of course..).

I think that this post sort of sums everything up.
 
What you posted is all irrelevant. AMD's strategy with Mantle is to get it into the game engines that power 80%+ of PC games, those engines being Frostbite 3, Cry Engine 3, and Unreal Engine 3/4. If they get Mantle into those game engines, 80-85% of game developers will automatically have full access to Mantle. Everyone using Frostbite 3 (and EA has 10 titles in development USING Frostbite 3) will have Mantle. The next generation Dragon Age: Inquisition game will have Mantle. (among other EA titles).

It doesn't matter if AMD talks to 5000 developers. They don't need to. They don't need to have the same API as the Xbox 1. They DONT need the same API as the PS4. That is completely irrelevant. They are attempting to get it into the game engines that power the majority of games, period, and if they do that they can have support in over 80% of multiplatform games.
 
No shit, thats pretty much exactly what I posted. Read the rest of this thread to see what I'm responding to.
 
No shit, thats pretty much exactly what I posted. Read the rest of this thread to see what I'm responding to.

I see! My apologies then. I'm certainly guilty of popping into a thread now and then without reading the full context of what is being discussed. My bad!
 
Back
Top