<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Yang, in Software Archive</title>
    <link>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988181#M27665</link>
    <description>&lt;P&gt;&lt;BR /&gt;Yang,&lt;/P&gt;
&lt;P&gt;The Intel IPP functions are threaded with OpenMP. The affinity setting in the IPP will also impact the IPP. Please check if the following environment setting could work for you: &lt;BR /&gt;KMP_AFFINITY=explicit,proclist=[1-6:1],granularity=fine&lt;/P&gt;
&lt;P&gt;You can check the following article regarding the these setting&lt;BR /&gt;&lt;A href="http://software.intel.com/sites/products/documentation/hpc/composerxe/en-us/2011Update/fortran/lin/optaps/common/optaps_openmp_thread_affinity.htm"&gt;http://software.intel.com/sites/products/documentation/hpc/composerxe/en-us/2011Update/fortran/lin/optaps/common/optaps_openmp_thread_affinity.htm&lt;/A&gt;&lt;/P&gt;
&lt;P&gt;Thanks,&lt;BR /&gt;Chao&lt;/P&gt;</description>
    <pubDate>Tue, 15 Jan 2013 02:34:19 GMT</pubDate>
    <dc:creator>Chao_Y_Intel</dc:creator>
    <dc:date>2013-01-15T02:34:19Z</dc:date>
    <item>
      <title>How to set thread affinity when using IPP library</title>
      <link>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988180#M27664</link>
      <description>&lt;P&gt;System detail:&lt;/P&gt;
&lt;P&gt;Computer we are using is an&amp;nbsp;Intel i7-2600. It&amp;nbsp;is 4 core CPU with hyper thread which give 8 virtual cores.&amp;nbsp;OS is Windows 7.&lt;/P&gt;
&lt;P&gt;Issue:&lt;/P&gt;
&lt;P&gt;We&amp;nbsp;have an image processing&amp;nbsp;process,&amp;nbsp;We like to run it in upper 6 cores and leave lower 2 cores for device control. I use setAffinity to set image processing process to upper 6 cores. To test the affinity, I use Windows task manager to monitor which core is used. It works as I expected that 6 cores busy when I use C++ code in my algorithm. However when I opptimize my code to use IPP library, it start to use all 8 cores; and ignore my Cpu affinity setting. Is there any way I can set CPU affinity to IPP library? Any help is appreciated.&lt;/P&gt;
&lt;P&gt;I have test ippSetAffinity. It does not make difference.&lt;/P&gt;
&lt;P&gt;Yang&lt;/P&gt;
&lt;P&gt;Abbott Point of Care&lt;/P&gt;</description>
      <pubDate>Fri, 11 Jan 2013 18:18:56 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988180#M27664</guid>
      <dc:creator>yangzhang</dc:creator>
      <dc:date>2013-01-11T18:18:56Z</dc:date>
    </item>
    <item>
      <title>Yang,</title>
      <link>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988181#M27665</link>
      <description>&lt;P&gt;&lt;BR /&gt;Yang,&lt;/P&gt;
&lt;P&gt;The Intel IPP functions are threaded with OpenMP. The affinity setting in the IPP will also impact the IPP. Please check if the following environment setting could work for you: &lt;BR /&gt;KMP_AFFINITY=explicit,proclist=[1-6:1],granularity=fine&lt;/P&gt;
&lt;P&gt;You can check the following article regarding the these setting&lt;BR /&gt;&lt;A href="http://software.intel.com/sites/products/documentation/hpc/composerxe/en-us/2011Update/fortran/lin/optaps/common/optaps_openmp_thread_affinity.htm"&gt;http://software.intel.com/sites/products/documentation/hpc/composerxe/en-us/2011Update/fortran/lin/optaps/common/optaps_openmp_thread_affinity.htm&lt;/A&gt;&lt;/P&gt;
&lt;P&gt;Thanks,&lt;BR /&gt;Chao&lt;/P&gt;</description>
      <pubDate>Tue, 15 Jan 2013 02:34:19 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988181#M27665</guid>
      <dc:creator>Chao_Y_Intel</dc:creator>
      <dc:date>2013-01-15T02:34:19Z</dc:date>
    </item>
    <item>
      <title>&gt;&gt;...However when I opptimize</title>
      <link>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988182#M27666</link>
      <description>&amp;gt;&amp;gt;...However when I opptimize my code to use IPP library, it start to use all 8 cores; and ignore my Cpu affinity setting...

This is expected result because OpenMP based processing in IPP library ( for some functions ) takes control of the thread management. However, if you get a list of all threads in the application using Win32 API you could set your own affinities for these threads. There is an example in MSDN on how to do this.</description>
      <pubDate>Sat, 19 Jan 2013 05:13:00 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Archive/How-to-set-thread-affinity-when-using-IPP-library/m-p/988182#M27666</guid>
      <dc:creator>SergeyKostrov</dc:creator>
      <dc:date>2013-01-19T05:13:00Z</dc:date>
    </item>
  </channel>
</rss>

