AMD Posts Mantle Whitepaper

by Ryan Smith on 5/29/2014 12:00 PM EST


Back to Article

  • nathanddrews - Thursday, May 29, 2014 - link

    I'm guessing that this was all part of a more elaborate plan.
    1. "Look at how evil NVIDIA are with their GameWorks!"
    2. "Look at how open we are about Mantle!"
  • nathanddrews - Thursday, May 29, 2014 - link

    In hindsight... not that elaborate. Reply
  • Wreckage - Thursday, May 29, 2014 - link

    The title should be "AMD posts more marketing material, Mantle still completely proprietary" Reply
  • CiccioB - Thursday, May 29, 2014 - link

    I wonder how one can doubt that a low level API can have better performances than a high level API. The main problem of developing something is not about performances. It is about how much of the market you can cover with the smaller effort.
    That's why high level API in all fields are used. Tha's why programs are written in C or bigger one in C++ instead of ASM.
    You can go as long as you want to demontrate that low level programming in ASM gives you more performances than using an high level language.
    The fact is that low level programming, in all fields (now even in embedded marked) is not profitable. You need a lot of efforts to maintain your product and to follow market evolution.

    About this very API, the problem is that is is linked tightly to GCN architecture. What will happen when a new architecture will be developed?
    On the other hand, how much can AMD change its architecture if they have an API that works only if some features/limitations of that one are respected?
    If you have a large program written in ASm you cannot use another microprocessor with a different ASM language, nor you can't even exploit a eventual upgrade of the same micro.
    Look at Motorola 68000 and Amiga development story to know what it leams being tied to a particular architecture.
  • CiccioB - Thursday, May 29, 2014 - link

    leams = means

    Still waiting for that edit button someday :D
  • aruisdante - Thursday, May 29, 2014 - link

    So... Mantle is definitely not Assembly-level low-level. It's still primarily a C/C++ construct. It just exposes a Mantle specific API instead of using the DirectX/OpenGL one, and it uses Mantle specific data structures in some cases. Reply
  • CiccioB - Friday, May 30, 2014 - link

    Well, not. It is a low level language that can be used only on one architecture. Set you the low level threshold, but it is like as you are using a custom LISP-like language for yopur customized MIPS and shows that you get better performance than using C++ that runs on anything in the world.
    You may have all the performance advantages you want, but you have long time to convience all the world to use both your MIPS implementation and your custom language which cannot evolve in future not to disrupt what that SW+HW tight integration brings today.

    This kind of APIs are a lock-in issue. 3Dfx's Glide were the same, and infact as soon ad DX started to be used widely they died miserably because they could not sustain their advantages using new architectures (32-bit depth color and T&L just to name some).
    Market compatibility, maintanability and easy porting to other/new/future architectures are factors that for complex projects are much more important that squeezing 10 more frame per second just for 25% of the market today.

    You may thing that doble implementation (Mantle+DX) can be the solution, but that is only if AMD pays for the extra work, otherwise none is ever thinking of using a limiting (market wise) API to create something that is just barely faster than using another that guarantess 100% market cover.
  • przemo_li - Friday, May 30, 2014 - link

    Funny You bring, DX.

    DX12 will bring many same architectural decisions.

    That mean at least MS see AMD architecture as mature enough for serving as base for next gen of GPU advancement.
    (DX12 is based on DX from console which evolved around AMD APUs capabilities)
  • pTmdfx - Saturday, May 31, 2014 - link

    Nah. Shader programs are still HLSL. Mantle just makes the application closer to the metal while still having a thin abstraction. It isn't even a "low-level language" but a paradigm of controlling the GPU. You sounds going too far. Yep. Reply
  • pTmdfx - Saturday, May 31, 2014 - link

    By the way, glad to tell you that cross-platform game engines do not use graphics API directly. They will wrap it into their own internal abstractions so that the engines can target numerous platforms with different APIs. So eventually there is nothing different from traditional APIs - just that the game developers have now more power to utilize the hardware to the edge, and the work of supporting it can be mostly hidden internally by the game engines. Reply
  • DanNeely - Thursday, May 29, 2014 - link

    The next major versions of Direct3d and OpenGL are incorporating many of the enhancements offered in Mantle. The key wins it offers are by changing out parts of the current standard rendering pipeline to reflect that the relative performance balance between the GPU and CPU has shifted by multiple orders of magnitude in favor of the former. The result of this is that what were cheap CPU steps a decade ago have become performance bottlenecks today.

    As a proprietary API Mantle's advantage is mostly in that it's able to hit the market much faster than new standards can. This gets AMD out in front, and with Mantle development being primarily pushed towards the developers of AAA game engines, the cost of adding support for the new standards in a few years will mostly fall on companies that can easily afford it and who're already willing to spend huge piles of cash to get every last sliver of performance, and already using every non-portable GPU family specific hack they're aware of in their low level code anyway (so supporting Mantle is business as usual for them).
  • TEAMSWITCHER - Thursday, May 29, 2014 - link

    I have to believe that AAA game makers are tired of adding support for platform specific API's - especially for only a sliver more performance on desktop PC's with graphic cards that don't have a performance problem to begin with. That and almost all mobile computing solutions now use Intel graphics. So Mantle is more pain than gain. Reply
  • Tetracycloide - Thursday, May 29, 2014 - link

    Platform specific? Didn't AMD do the hardware for both the 360 and the PS4? I don't think there's much crossover in AAA development from console/desktop to mobile. I mean yes, there have been some ports back and forth but it's pretty uncommon. Porting from a console to another console or to the PC is the norm for almost everything though. Reply
  • tobi1449 - Thursday, May 29, 2014 - link

    Yes, but: None of the consoles support Mantle. They each feature their respective companies low level libraries. Reply
  • przemo_li - Friday, May 30, 2014 - link

    We do not know (and wont as XOne/PS4 APIs are proprietary), how much influence MS/Sony had over them. They could be slightly or significantly different.

    Also XOne sports Win, PS4 have OS that is "embeded", while Mantle is "just" driver.
  • kyuu - Thursday, May 29, 2014 - link

    You "have to believe" that because you don't seem to understand just how much of a bottleneck CPU-bound operations are (the big one being draw calls). It's not about eeking "a sliver more performance" from desktop GPUs. The issue of inadequate CPU performance (relative to GPU performance) is holding back developers who want to push the level of graphics fidelity in their games. That's why CIG is implementing Mantle-support for Star Citizen, for example. Reply
  • przemo_li - Friday, May 30, 2014 - link

    Where did You heard that?


    (Sad) story of the industry is that You need many backends. PS3 have its API, PS4 have its API, X360 have its API, XOne have its API, Win7 have its API, Win8 have its API, Linux/OSX have smae api but with different feature level supported. Android/iOS have their API, WinP have its API.

    You know where I'm going....
  • killeak - Friday, May 30, 2014 - link

    Is true that Direct3D 12 will be very similar to Mantle, but I really doubt that about OpenGL Next will be like that.

    While MS can rewrite the API from scratch, and they did that several times in the DX history, OpenGL is still based on the old state machine API from the '92. I don't see how Khronos can make an OpenGL that look like mantle or direct3D 12, without breaking compatibility and redesigning almost everything. They never did it, not with GL 2.0 nor 3 nor 4.

    So, no I won't have much hope in OpenGL 5 to be like Mantle or Direct3D 12. To be honest, OpenGL has it's place. Is a high level API that works in many different hardware in many different OS. We need an API like that for many things from small games to CAD apps, is just not ideal for big AAA games...
  • przemo_li - Friday, May 30, 2014 - link

    Duno about OpenGL.

    Khronos did not shared any such thoughts. And current ARB extensions lack in many palces compared to Mantle (memory management, parallelism, acces to engines on GPU, comand buffer construction).

    MS seam to follow on "general" terms. But they do not talk much about specifics to see how far apart or how similar concepts will be in both of them.
  • haikuginger - Thursday, May 29, 2014 - link

    Given that both major console players are using AMD hardware this generation, Mantle makes a lot of sense. It'd be a tough sell to get developers to write for it if it was for PC hardware only, but given that developers are already going to be optimizing for the AMD hardware in the Xbox One and the PS4, it's not that much of a stretch to apply the same optimizations to the PC space. AMD hardware in game consoles makes platform-specific code much more general-purpose. Reply
  • KonradK - Thursday, May 29, 2014 - link

    A lower level API does not need to be hardware specific API.
    A lower level API is the API less abstract than a higher level API.
    Level of Direct 3D is lower than level of GDI, but Direct 3D is not hardware specific.

  • lefty2 - Thursday, May 29, 2014 - link

    Pretty old news. That white paper was actually posted 3 months ago. Reply
  • gostan - Saturday, May 31, 2014 - link

    You are wrong. This version is newer because they have renamed the file. There will be another version coming out in 2015. No changes in content but it will be treated as a new version because they will rename it once again.

    In 2016, AMD will still be TALKING about Mantle. It will be like their APU/Fusion. It's so awesome, device makers have decided not to take it seriously. The core business of AMD is about having a good conversion, hence their motto - Let's Talk!
  • MadMinstrel - Thursday, May 29, 2014 - link

    "Among the subjects covered are Mantle's memory model, execution model, pipeline model, and the basic tenants of where low-level APIs can reduce overhead and improve performance over high level APIs."

    Tenets. Unless there's gremlins in the code.
  • Krysto - Friday, May 30, 2014 - link

    Oh good. Now Microsoft can clone it and call it "the totally different DirectX 12". Reply

Log in

Don't have an account? Sign up now