<?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 Multi-threaded PSSYEVD not threading actually in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142692#M26440</link>
    <description>&lt;P&gt;Dear all,&lt;/P&gt;&lt;P&gt;When calling the SCALAPACK routine PSSYEVD with several threads, only the first MPI task is actually threading as expected.&lt;/P&gt;&lt;P&gt;I have met this issue on both my laptop and on a supercomputer, both running with MKL 2019.2.&lt;/P&gt;&lt;P&gt;In practice, the "top" command on my laptop when running with 2 MPI tasks and 2 MKL threads&amp;nbsp;shows that&amp;nbsp;the first MPI task runs at 200%, whereas the second one runs at 100%.&lt;/P&gt;&lt;P&gt;Note that this problem disappear when I call the double precision routine PDSYEVD instead or when I call the more regular&amp;nbsp;PSSYEV.&lt;/P&gt;&lt;P&gt;Best,&lt;/P&gt;&lt;P&gt;Fabien&lt;/P&gt;</description>
    <pubDate>Thu, 09 May 2019 10:05:53 GMT</pubDate>
    <dc:creator>fabien_b_</dc:creator>
    <dc:date>2019-05-09T10:05:53Z</dc:date>
    <item>
      <title>Multi-threaded PSSYEVD not threading actually</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142692#M26440</link>
      <description>&lt;P&gt;Dear all,&lt;/P&gt;&lt;P&gt;When calling the SCALAPACK routine PSSYEVD with several threads, only the first MPI task is actually threading as expected.&lt;/P&gt;&lt;P&gt;I have met this issue on both my laptop and on a supercomputer, both running with MKL 2019.2.&lt;/P&gt;&lt;P&gt;In practice, the "top" command on my laptop when running with 2 MPI tasks and 2 MKL threads&amp;nbsp;shows that&amp;nbsp;the first MPI task runs at 200%, whereas the second one runs at 100%.&lt;/P&gt;&lt;P&gt;Note that this problem disappear when I call the double precision routine PDSYEVD instead or when I call the more regular&amp;nbsp;PSSYEV.&lt;/P&gt;&lt;P&gt;Best,&lt;/P&gt;&lt;P&gt;Fabien&lt;/P&gt;</description>
      <pubDate>Thu, 09 May 2019 10:05:53 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142692#M26440</guid>
      <dc:creator>fabien_b_</dc:creator>
      <dc:date>2019-05-09T10:05:53Z</dc:date>
    </item>
    <item>
      <title>actually this looks very</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142693#M26441</link>
      <description>&lt;P&gt;actually this looks very strange because of our implementation is the same for single and double precision data types.&amp;nbsp;&lt;/P&gt;&lt;P&gt;Am I understand you right that&amp;nbsp;you run&amp;nbsp;PSSYEVD executable on SMP machine and expect to see different threading based on top output?&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Thu, 09 May 2019 11:54:39 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142693#M26441</guid>
      <dc:creator>Gennady_F_Intel</dc:creator>
      <dc:date>2019-05-09T11:54:39Z</dc:date>
    </item>
    <item>
      <title>Hi,</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142694#M26442</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;Let me be more precise. Here is why I do on my 4-core laptop:&amp;nbsp;&lt;/P&gt;
&lt;PRE class="brush:bash; class-name:dark;"&gt;&amp;gt; export OMP_NUM_THREADS=2
&amp;gt; export MKL_NUM_THREADS=2
&amp;gt; mpirun -np 2 ./a.out &amp;amp;
&amp;gt; top
  PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND                                                                                                                                       
15075 bruneval  20   0 2883124 422376  23480 R 198.7   2.6   0:21.94 a.out                                                                                                                                         
15076 bruneval  20   0 1236508 237804  18008 R  99.7   1.5   0:11.64 a.out   &lt;/PRE&gt;

&lt;PRE class="brush:bash; class-name:dark;"&gt;
&amp;nbsp;&lt;/PRE&gt;

&lt;P&gt;The top shows two unbalanced MPI tasks. This is not what I expect. PDSYEVD would produce two tasks with 200% each.&lt;/P&gt;
&lt;P&gt;Could someone reproduce this?&lt;/P&gt;
&lt;P&gt;Best,&lt;/P&gt;
&lt;P&gt;Fabien&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Thu, 09 May 2019 12:10:00 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Multi-threaded-PSSYEVD-not-threading-actually/m-p/1142694#M26442</guid>
      <dc:creator>fabien_b_</dc:creator>
      <dc:date>2019-05-09T12:10:00Z</dc:date>
    </item>
  </channel>
</rss>

