The AMD Mantle Thread

[VulgarDisplay]

Isn't thief 4 on Unreal 3? It's getting mantle. You know it's possible.

Like you said, if it doesn't come to Unreal it's purely politics/sponsorship related.

Can you imagine a world where AMD and Nvidia traded Mantle and Phsyx between them straight up. Games would overnight become much more interesting, and I imagine all that extra draw call use allowed by mantle could really benefit Physx performance.

 

[Phynaz]

You'll get similar performance with any recent CPU. It's possible that Kaveri might have extra performance if DICE has used the extra shaders in asymmetric crossfire - but even then it would be 10% maximum and more like 5% I think.

Kaveri is out, where's Mantel?

 

[moonbogg]

Kaveri is out, where's Mantel?

Honestly, if they said January then I'd expect it as late in January as possible. Like midnight on the last day or something.

 

[Carfax83]

Isn't thief 4 on Unreal 3? It's getting mantle. You know it's possible.

Like you said, if it doesn't come to Unreal it's purely politics/sponsorship related.

Unreal Engine is so customizable that any developer could incorporate Mantle I would say in their game, like what Eidos Montreal is doing..

But that doesn't mean EPIC will integrate Mantle deep into the engine, like they did PhysX, or like DICE has done with FB 3.

Anyway, I'm kinda disappointed that Thief 4 is using Unreal Engine 3. That engine needs to die.. It's a great engine but it's so old!

 

[OCGuy]

Honestly, if they said January then I'd expect it as late in January as possible. Like midnight on the last day or something.

Latest slide said Q1. My guess is 2nd week in Feb.

 

[Paul98]

Kaveri is out, where's Mantel?

Out, developers are working with it. Now where are games with mantle, waiting on BF4 to not be broken.

 

[OCGuy]

Out, developers are working with it. Now where are games with mantle, waiting on BF4 to not be broken.

I play BF4 regularly....if DX works for me, then certainly Mantle can.

 

[blastingcap]

"How one engine is shaping the future of EA Games"
http://www.engadget.com/2013/11/19/electronic-arts-frostbite-battlefield-mass-effect/

(warning: marketing like material )

This is a tangent from the article, but I wanted to float something out there that I first floated a few years ago: why not build a megalibrary of ultra-high rez textures? For instance you could scan an oak tree to an extremely high resolution and make it the "standard" oak texture. Do the same with gold. Etc. If you don't need that kind of resolution that's fine, downsample it.

Every game dev could then draw from this massive open-source super-high texture library and save money, so that more money goes into other aspects of game development. I guess the main problem is free-ridership?

 

[SiliconWars]

Kaveri is out, where's Mantel?

I dunno, what's "Mantel"?

Next time just report the posts instead of responding.

-Rvenger

 

[Spjut]

I expect Mantle to die, but not before throwing a scare into Microsoft so that DX12 is more Mantle-like. And that's fine by me because it's a victory for all gamers.

The possible negative with DX12 is that it likely would require the newest Windows, which will greatly slow down its adoption rate. Just look at how long DX9 lived due to XP.
Microsoft has already made DX11.1/11.2 exclusive for Windows 8.x
I'd guess the next OpenGL could be supported on today's OSes though

 

[SiliconWars]

http://www.pcgamesn.com/amds-mantle-api-how-it-works-and-why-pc-developers-will-adopt-it

Hallock admitted, “The clever thing about Mantle is that it’s very similar to the types of code these game developers are already creating for consoles, so the process of porting is quite simple for them.”

 

[Gloomy]

Ah, the good old "push in-pull out." That's a classic. AMD does know what the market loves. In my opinion it's a good move for someone coming in from behind.

But Nvidia knows how to reach around and give a more balanced experience, so they're going to have to work hard to come out on top.

:hmm:

 

[Slomo4shO]

http://www.pcgamesn.com/amds-mantle-api-how-it-works-and-why-pc-developers-will-adopt-it

How about some developer prospective instead of the AMD mouthpiece throwing out marketing talking points...

 

[blastingcap]

Ah, the good old "push in-pull out." That's a classic. AMD does know what the market loves. In my opinion it's a good move for someone coming in from behind.

But Nvidia knows how to reach around and give a more balanced experience, so they're going to have to work hard to come out on top.

:hmm:

I thought I was at a hardware forum, not a softcore pornum...

 

[smackababy]

How about some developer prospective instead of the AMD mouthpiece throwing out marketing talking points...

I would prefer an actual "This is code to do X on the PS4 API and this is the code to the same thing on Mantle; conversely, this is "worse" code doing the same thing on DX." That way, all this "the code is similar" garbage can be put to rest.

Of course, something so simple would be too simple to debunk the rumor of it being a copy paste job from the console API that AMD wants you to think in their slides.


Another thing I will say is that it is rather telling there hasn't been any leaks of such code samples. Even anonymously leaked code would be intriguing for people not sippin' the Kool Aid hoping it contains a //Titan's mom is so fat her shader pipelane is over 9000mb/s*

*my attempt at "ridiculing" Titan. =( Sorry.

 

[VulgarDisplay]

I would prefer an actual "This is code to do X on the PS4 API and this is the code to the same thing on Mantle; conversely, this is "worse" code doing the same thing on DX." That way, all this "the code is similar" garbage can be put to rest.

Of course, something so simple would be too simple to debunk the rumor of it being a copy paste job from the console API that AMD wants you to think in their slides.


Another thing I will say is that it is rather telling there hasn't been any leaks of such code samples. Even anonymously leaked code would be intriguing for people not sippin' the Kool Aid hoping it contains a //Titan's mom is so fat her shader pipelane is over 9000mb/s*

*my attempt at "ridiculing" Titan. =( Sorry.

Just curios how often you gain access to alpha software source code? There is so much misinformation being spread About mantle that some things that were pure speculation on the part of media and forums are now being called amd lies.

AMD never said mantle is the same as console. That started on anandtechs first mantle article as pure speculation.

 

[TerryMathews]

I would prefer an actual "This is code to do X on the PS4 API and this is the code to the same thing on Mantle; conversely, this is "worse" code doing the same thing on DX." That way, all this "the code is similar" garbage can be put to rest.

Wouldn't a software company consider those routines to be confidential and proprietary trade secrets?

 

[krumme]

That was a strange article...

Reads like pure frostbite marketing, and ends with you can't have it unless you're with EA.

Very strange.

Yeaa
But its a good description how the dev studio community is evolving inside EA.

When we talk mantle i think we forget its also pretty much a consequence of going from an engine for each studio to an engine shared for all.

So instead of 40 brains they have 2000 brains. With all that power obviously you can adopt a lot of tech. And a part of that is eg using mantle.

The fragmentation of engines is comming to an end but perhaps we will see more fragmentation of the tecynollgy used like eg. Low end api?

 

[krumme]

This is a tangent from the article, but I wanted to float something out there that I first floated a few years ago: why not build a megalibrary of ultra-high rez textures? For instance you could scan an oak tree to an extremely high resolution and make it the "standard" oak texture. Do the same with gold. Etc. If you don't need that kind of resolution that's fine, downsample it.

Every game dev could then draw from this massive open-source super-high texture library and save money, so that more money goes into other aspects of game development. I guess the main problem is free-ridership?

Its a good thought.

And the effect you want is pretty much what EA tries to get imho with this betting on one engine fb3.

The problem is ofcourse you want to have a "leg up" with the competition
so you want to share
...
And you dont want to share lol

 

[Carfax83]

Here's a thought of how Microsoft could lower the next version of Direct3D's draw call overhead:

We also took the opportunity to go and highly customise the command processor on the GPU. Again concentrating on CPU performance... The command processor block's interface is a very key component in making the CPU overhead of graphics quite efficient. We know the AMD architecture pretty well - we had AMD graphics on the Xbox 360 and there were a number of features we used there. We had features like pre-compiled command buffers where developers would go and pre-build a lot of their states at the object level where they would [simply] say, "run this". We implemented it on Xbox 360 and had a whole lot of ideas on how to make that more efficient [and with] a cleaner API, so we took that opportunity with Xbox One and with our customised command processor we've created extensions on top of D3D which fit very nicely into the D3D model and this is something that we'd like to integrate back into mainline 3D on the PC too - this small, very low-level, very efficient object-orientated submission of your draw [and state] commands.

Source

Looks like they may be going the extension route, similar to what OpenGL did with their bindless extensions.

 

[krumme]

I would prefer an actual "This is code to do X on the PS4 API and this is the code to the same thing on Mantle; conversely, this is "worse" code doing the same thing on DX." That way, all this "the code is similar" garbage can be put to rest.

Of course, something so simple would be too simple to debunk the rumor of it being a copy paste job from the console API that AMD wants you to think in their slides.


Another thing I will say is that it is rather telling there hasn't been any leaks of such code samples. Even anonymously leaked code would be intriguing for people not sippin' the Kool Aid hoping it contains a //Titan's mom is so fat her shader pipelane is over 9000mb/s*

*my attempt at "ridiculing" Titan. =( Sorry.

I am not here to smackyourpony but will just remind you that differences be it mantle or physx is not measured in bits but in $

It doesn't matter if 90% is directly portable if the 10% missing is highly difficult or the hidden cost creeps in. What matters is total cost of building the tech into the engine and applying it. I understand the perspective from floor programmers side but there is much more to it from eg management to user benefit.

 

[3DVagabond]

This is a tangent from the article, but I wanted to float something out there that I first floated a few years ago: why not build a megalibrary of ultra-high rez textures? For instance you could scan an oak tree to an extremely high resolution and make it the "standard" oak texture. Do the same with gold. Etc. If you don't need that kind of resolution that's fine, downsample it.

Every game dev could then draw from this massive open-source super-high texture library and save money, so that more money goes into other aspects of game development. I guess the main problem is free-ridership?

Artists (modelers) maintain their own libraries of textures. Some are as you said, photos. Some are "hand drawn" in PS or some other program. Some are procedural shaders. Then there's combinations of the above. But it's generally the modelers that maintain their own library.

 

[Will Robinson]

Blasting Cap wrote:

I expect Mantle to die, but not before throwing a scare into Microsoft so that DX12 is more Mantle-like. And that's fine by me because it's a victory for all gamers. __________________

Why would Mantle die?
It is designed for use on CGN hardware....designed and made by AMD.
You think NV and AMD will hand over their latest hardware to Microsoft and then tell them to design an API to work at close to the metal level with it?
I dont think so.
I think DX will remain as the "layer above" to ensure decent compatibility across all platforms with specific low level APIs to come...first with Mantle for CGN and then an in house API for NV architecture.

 

[BallaTheFeared]

I really hope not.

 

[Will Robinson]

How about some developer prospective instead of the AMD mouthpiece throwing out marketing talking points...

Wow...that was uncalled for.:thumbsdown: