<?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 Re:Mapping and pinning MPI tasks and threads in Intel® MPI Library</title>
    <link>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1260935#M7870</link>
    <description>&lt;P&gt;Hi Micheal,&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;We haven't heard back from you.&lt;/P&gt;&lt;P&gt;Let us if the given command satisfies your pinning requirements?&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;&lt;P&gt;Prasanth&lt;/P&gt;&lt;BR /&gt;</description>
    <pubDate>Wed, 03 Mar 2021 08:11:34 GMT</pubDate>
    <dc:creator>PrasanthD_intel</dc:creator>
    <dc:date>2021-03-03T08:11:34Z</dc:date>
    <item>
      <title>Mapping and pinning MPI tasks and OpenMP threads</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1257172#M7800</link>
      <description>&lt;P&gt;I am very confused with the Intel MPI and OpenMP mapping and I was hopping someone could help me to understand. I'll describe my problem with an example.&lt;/P&gt;
&lt;P&gt;I want to run an application (LAMMPS) which uses MPI and OpenMP.&amp;nbsp; The way that OpenMP is used here is via SMT and through the LAMMPS Intel package. The target architecture (compute node) is a 2x socket, 32 core x&amp;nbsp; socket, 2 SMT x core AMD chips.&lt;/P&gt;
&lt;P&gt;&lt;STRONG&gt;* Full node (64 MPI tasks, 2 OMP threads)&lt;/STRONG&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;LI-CODE lang="markup"&gt;export OMP_NUM_THREAD=2
mpirun -np 64 -pk intel 0 omp 2 -sf intel ./lmp &lt;/LI-CODE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I think the above is correct, but how can I achieve the following:&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&lt;STRONG&gt;* Half node ( split into 2 sockets, 16 each)&lt;/STRONG&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;LI-CODE lang="markup"&gt;export OMP_NUM_THREADS=2
mpirun -np 32 -pk intel 0 omp 2 -sf intel ...&lt;/LI-CODE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&lt;STRONG&gt;* Half node (all in 1 socket)&lt;/STRONG&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;LI-CODE lang="markup"&gt;export OMP_NUM_THREADS=2
mpirun -np 32 -pk intel 0 omp 2 -sf intel ./lmp &lt;/LI-CODE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I have been playing with the following variables, but I have not yet been able to do it:&lt;/P&gt;
&lt;LI-CODE lang="markup"&gt;I_MPI_PIN_DOMAIN
I_MPI_PIN_PROCESSOR_EXCLUDE_LIST
I_MPI_PIN_PROCESSOR_LIST
KMP_AFFINITY&lt;/LI-CODE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Thu, 18 Feb 2021 12:07:32 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1257172#M7800</guid>
      <dc:creator>cristmoltisanti</dc:creator>
      <dc:date>2021-02-18T12:07:32Z</dc:date>
    </item>
    <item>
      <title>Re: Mapping and pinning MPI tasks and OpenMP threads</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1258821#M7832</link>
      <description>&lt;P&gt;Hi Micheal,&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I am sorry for the late reply.&lt;/P&gt;
&lt;P&gt;As you have mentioned that you wanted to distribute the MPI ranks based on sockets.&lt;/P&gt;
&lt;P&gt;As per your two use cases :&lt;/P&gt;
&lt;P&gt;I) Half node ( split into 2 sockets, 16 each)&lt;/P&gt;
&lt;P&gt;In this case, as you want to distribute 16 ranks on each socket by launching a total of 32 ranks. There is no need to do pinning as this the default behavior.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;2)Half node (all in 1 socket)&lt;/P&gt;
&lt;P&gt;To launch all threads in a single socket try this command&lt;/P&gt;
&lt;P&gt;I_MPI_PIN_DOMAIN=socket:compact I_MPI_DEBUG=10 OMP_NUM_THREADS=2 mpirun -n 32 -ppn 2&amp;nbsp;./a.out&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;For more information on how to use I_MPI_PIN_DOMAIN, please refer (&lt;A href="https://software.intel.com/content/www/us/en/develop/documentation/mpi-developer-reference-linux/top/environment-variable-reference/process-pinning/interoperability-with-openmp.html" target="_blank" rel="noopener noreferrer"&gt;Interoperability with OpenMP* API (intel.com)&lt;/A&gt;)&lt;/P&gt;
&lt;P&gt;Let us know if you face problems.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Regards&lt;/P&gt;
&lt;P&gt;Prasanth&lt;/P&gt;</description>
      <pubDate>Wed, 24 Feb 2021 12:28:16 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1258821#M7832</guid>
      <dc:creator>PrasanthD_intel</dc:creator>
      <dc:date>2021-02-24T12:28:16Z</dc:date>
    </item>
    <item>
      <title>Re:Mapping and pinning MPI tasks and threads</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1260935#M7870</link>
      <description>&lt;P&gt;Hi Micheal,&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;We haven't heard back from you.&lt;/P&gt;&lt;P&gt;Let us if the given command satisfies your pinning requirements?&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;&lt;P&gt;Prasanth&lt;/P&gt;&lt;BR /&gt;</description>
      <pubDate>Wed, 03 Mar 2021 08:11:34 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1260935#M7870</guid>
      <dc:creator>PrasanthD_intel</dc:creator>
      <dc:date>2021-03-03T08:11:34Z</dc:date>
    </item>
    <item>
      <title>Re:Mapping and pinning MPI tasks and threads</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1262628#M7893</link>
      <description>&lt;P&gt;Hi Micheal,&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;We are closing this thread assuming your issue has been resolved. We will no longer respond to this thread. If you require additional assistance from Intel, please start a new thread. Any further interaction in this thread will be considered community only.&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;P&gt;Regards&lt;/P&gt;&lt;P&gt;Prasanth&lt;/P&gt;&lt;BR /&gt;</description>
      <pubDate>Tue, 09 Mar 2021 07:26:46 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Mapping-and-pinning-MPI-tasks-and-OpenMP-threads/m-p/1262628#M7893</guid>
      <dc:creator>PrasanthD_intel</dc:creator>
      <dc:date>2021-03-09T07:26:46Z</dc:date>
    </item>
  </channel>
</rss>

