官术网_书友最值得收藏!

Installing and configuring ATI video card drivers

In this recipe, we'll go into the details for installing and configuring the ATI video card drivers followed by the AMD Accelerated Parallel Processing (APP) SDK, OpenCL, and CAL++. Taking advantage of the ATI Stream technology, we can run computationally-intensive tasks—typically running on the CPU—that perform more quickly and efficiently. For more detailed information regarding the ATI Stream technology, visit www.amd.com/stream.

Getting ready

An Internet connection is required to complete this recipe. The preparation of kernel headers is also needed before starting this task, which is explained in the Preparing kernel headers recipe at the beginning of this chapter.

How to do it...

Let's begin installing and configuring the ATI drivers:

  1. Download the ATI display driver required for your system:
    cd /tmp/
    wget http://www2.ati.com/drivers/linux/amd-driver-installer-12-1-x86.x86_64.run
    

    We can also download the display driver from the following site: http://support.amd.com/us/gpudownload/Pages/index.aspx.

  2. Start the installation by typing the following command:
    sh amd-driver-installer-12-1-x86.x86_64.run
    
  3. When the setup completes, reboot your system for the changes to take effect and to prevent system instability.
  4. Install the dependencies needed for further steps:
    apt-get install libroot-python-dev libboost-python-dev libboost1.40-all-dev cmake
    
  5. Download and untar the AMD APP SDK according to your CPU architecture:
    wget http://developer.amd.com/Downloads/AMD-APP-SDK-v2.6-lnx64.tgz
    mkdir AMD-APP-SDK-v2.6-lnx64
    tar zxvf AMD-APP-SDK-v2.6-lnx64.tgz –C /tmp/AMD-APP-SDK-v2.6-lnx64
    cd AMD-APP-SDK-v2.6-lnx64
    
  6. Install the AMD APP SDK by issuing the following command:
    sh Install-AMD-APP.sh
    
  7. Set the ATI Stream paths in the .bashrc file:
    echo export ATISTREAMSDKROOT=/opt/AMDAPP/ >> ~/.bashrc
    source ~/.bashrc
    
  8. Download and compile calpp:
    cd /tmp/
    svn co https://calpp.svn.sourceforge.net/svnroot/calpp calpp
    cd calpp/trunk
    cmake .
    make
    make install
    
  9. Download and compile pyrit:
    cd /tmp/
    svn co http://pyrit.googlecode.com/svn/trunk/ pyrit_src
    cd pyrit_src/pyrit
    python setup.py build
    python setup.py install
    
  10. Build and install OpenCL:
    cd /tmp/pyrit_src/cpyrit_opencl
    python setup.py build
    python setup.py install
    
  11. Make a few changes to the cpyrit_calpp setup:
    cd /tmp/pyrit_source/cpyrit_calpp
    vi setup.py
    

    Look at the following line:

    VERSION = '0.4.0-dev'

    Replace it with:

    VERSION = '0.4.1-dev'

    Also, look at the following line:

    CALPP_INC_DIRS.append(os.path.join(CALPP_INC_DIR, 'include'))

    Replace it with:

    CALPP_INC_DIRS.append(os.path.join(CALPP_INC_DIR, 'include/CAL'))
  12. Finally, add the ATI GPU module to pyrit:
    python setup.py build
    python setup.py install
    
Note

To show the available CAL++ devices and CPU cores, we issue the following command:

pyrit list_cores

To perform a benchmark, we simply type:

pyrit benchmark
主站蜘蛛池模板: 铜川市| 丹阳市| 南召县| 驻马店市| 平原县| 全椒县| 东至县| 潍坊市| 文登市| 枣阳市| 平远县| 南澳县| 宾川县| 桑日县| 金寨县| 襄樊市| 余干县| 海南省| 东乡族自治县| 龙江县| 河东区| 抚顺市| 黄石市| 乐清市| 麻城市| 富源县| 巴中市| 时尚| 和林格尔县| 拉萨市| 思茅市| 工布江达县| 上栗县| 亚东县| 西青区| 图们市| 凌海市| 大港区| 华阴市| 思南县| 万年县|