<?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 threads of mkl in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805751#M3407</link>
    <description>&lt;DIV id="tiny_quote"&gt;&lt;DIV style="margin-left: 2px; margin-right: 2px;"&gt;Quoting &lt;A jquery1297173351642="63" rel="/en-us/services/profile/quick_profile.php?is_paid=&amp;amp;user_id=508582" href="https://community.intel.com/en-us/profile/508582/" class="basic"&gt;vadim_l&lt;/A&gt;&lt;/DIV&gt;&lt;DIV style="background-color: #e5e5e5; margin-left: 2px; margin-right: 2px; border: 1px inset; padding: 5px;"&gt;&lt;I&gt;Tell me please on how many physical cores MKL shows significant performance?&lt;/I&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;P&gt;&lt;/P&gt;Hi,&lt;BR /&gt;&lt;BR /&gt;In general it depends on MKL domain and tasks you are using. For BLAS for example the best performance will be when used 1-cpu for core (similar toHT is off).&lt;BR /&gt;But for other parts you can get some improvements in using HT.</description>
    <pubDate>Tue, 08 Feb 2011 14:00:24 GMT</pubDate>
    <dc:creator>barragan_villanueva_</dc:creator>
    <dc:date>2011-02-08T14:00:24Z</dc:date>
    <item>
      <title>threads of mkl</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805748#M3404</link>
      <description>Hello,&lt;BR /&gt;I have a question for the library mkl&lt;BR /&gt;I have intel core i3 with 4 logical cores, and was installed Parallel Studio 2011 and the intel composer XE 2011 with version mkl 10.3.1. When I create a project in visual studio I chose the menu option - 'parallel' mkl in composer ,but when the program was started, windows task manager show only 2 working cores, what's the problem?&lt;BR /&gt;I tried using the function mkl_set_num_threads(4),but nothing changed.</description>
      <pubDate>Tue, 08 Feb 2011 06:38:37 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805748#M3404</guid>
      <dc:creator>vadim_l</dc:creator>
      <dc:date>2011-02-08T06:38:37Z</dc:date>
    </item>
    <item>
      <title>threads of mkl</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805749#M3405</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;Maybe it's because MKL recognizes HT and limits threads to 1 per core by default to get the best performance.&lt;BR /&gt;&lt;BR /&gt;To allow using all logical cores please set environment MKL_DYNAMIC=false together with settting MKL_NUM_THREADS=4 or using the function mkl_set_num_threads(4).</description>
      <pubDate>Tue, 08 Feb 2011 06:50:40 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805749#M3405</guid>
      <dc:creator>barragan_villanueva_</dc:creator>
      <dc:date>2011-02-08T06:50:40Z</dc:date>
    </item>
    <item>
      <title>threads of mkl</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805750#M3406</link>
      <description>Thanx, Victor,&lt;BR /&gt;the function mkl_set_num_threads(4) together with omp_set_dynamic(false) helped me. However, the performance is no much increased.&lt;BR /&gt;Tell me please on how many physical cores MKL shows significant performance?</description>
      <pubDate>Tue, 08 Feb 2011 11:35:01 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805750#M3406</guid>
      <dc:creator>vadim_l</dc:creator>
      <dc:date>2011-02-08T11:35:01Z</dc:date>
    </item>
    <item>
      <title>threads of mkl</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805751#M3407</link>
      <description>&lt;DIV id="tiny_quote"&gt;&lt;DIV style="margin-left: 2px; margin-right: 2px;"&gt;Quoting &lt;A jquery1297173351642="63" rel="/en-us/services/profile/quick_profile.php?is_paid=&amp;amp;user_id=508582" href="https://community.intel.com/en-us/profile/508582/" class="basic"&gt;vadim_l&lt;/A&gt;&lt;/DIV&gt;&lt;DIV style="background-color: #e5e5e5; margin-left: 2px; margin-right: 2px; border: 1px inset; padding: 5px;"&gt;&lt;I&gt;Tell me please on how many physical cores MKL shows significant performance?&lt;/I&gt;&lt;/DIV&gt;&lt;/DIV&gt;&lt;P&gt;&lt;/P&gt;Hi,&lt;BR /&gt;&lt;BR /&gt;In general it depends on MKL domain and tasks you are using. For BLAS for example the best performance will be when used 1-cpu for core (similar toHT is off).&lt;BR /&gt;But for other parts you can get some improvements in using HT.</description>
      <pubDate>Tue, 08 Feb 2011 14:00:24 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/threads-of-mkl/m-p/805751#M3407</guid>
      <dc:creator>barragan_villanueva_</dc:creator>
      <dc:date>2011-02-08T14:00:24Z</dc:date>
    </item>
  </channel>
</rss>

