What’s New in AMD APP (formerly ATI Stream) SDK v2.3


AMD recently released version 2.3 of the AMD Accelerated Parallel Processing (formerly ATI Stream) SDK with the following new features:

  • Improved OpenCL™ runtime performance:
    • Improved kernel launch times
    • Improved PCIe transfer times
    • Enabled DRMDMA for the ATI Radeon 5000 Series and AMD Radeon 6800 GPUs that are specified in the Supported Devices
  • Increased size of staging buffers
  • Enhanced Binary Image Format (BIF)
  • Support for UVD video hardware component through OpenCL (Windows® 7)
  • Support for AMD Radeon™ HD 6900 and HD 6800 Series graphics
  • Support for AMD E-Series, C-Series, and G-Series platforms
  • Support for OpenCL math libraries: FFT and BLAS-3

Included are previews of the following features:

  • Preview Feature: An optimization pragma for unrolling loops.
  • Preview Feature: Support for CPU/x86 image. This enables the support for Image formats, as described in the Khronos specification for OpenCL, to be run on the x86 CPU. It is enabled by the following environment variable in your application: CPU_IMAGE_SUPPORT.

In conjunction, the AMD APP (formerly ATI Stream) Profiler v2.1 and AMD APP (formerly Stream) KernelAnalyzer (SKA) v1.7 were released with a number of enhancements to help make programming with the SDK even easier.

Download the SDK today and let us know what you think.

OpenCL is a trademark of Apple Inc., and is used under license by Khronos.

This post is the opinion of the author and may not represent AMD’s positions, strategies or opinions. Links to third party sites and references to third party trademarks are provided for convenience and illustrative purposes only. Unless explicitly stated, AMD is not responsible for the contents of such links, and no third party endorsement of AMD or any of its products is implied.

6 Responses

  1. Tsiolkovsky

    When will the ATI Stream Profiler for Linux be available as an application with a GUI too? BTW shouldn’t AMD be using some cross-platform GUI toolkit like Qt so that all tools could right away be available on Linux also.

  2. martin

    yea nice we wait for gpu physics like havok bullet in 4000 5000 6000 series
    writeen in directcompute 4 or 5 10.1 11 dx

    and start take benefit of it with some software like winrar compression kgb …
    and of course especially in games like ai physics ….

  3. Datsun

    So, AMD is beeing prepared for an intangible processor while AMD has a vision of computing beyond silicon or any semiconductor materials. That is why AMD has detached its manufacturing ARM.

    By using quantum computing mechanics, processor circuits can be made dynamically like an FPGA. A Blender like machine with a cylinder which is composed of multiple particle ray sources. When the machine is rotating, each particle rays will dynamically self aligned to make a logic circuit. The blender has a magnetic resonance that generates magnetic fields. The result is a holographic tube which is a replacement for any processors. The core of tube will process scalar based data while the layers to periphery will process vectors data.

  4. Frank Mueller

    No Linux support for E350/Zacate Radeon HD 6310 yet for the the SDK? I can’t find a driver. Sad.