<?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 I believe the default value in Intel® MPI Library</title>
    <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038279#M4264</link>
    <description>&lt;P&gt;I believe the default value is actually 'compact'.&amp;nbsp; That means the first 5 consecutive ranks will be pinned to the first socket, and the next set of 5 will be pinning to the second socket.&lt;/P&gt;

&lt;P&gt;If you want to do 'scatter' instead (which, in your case, means alternating ranks will be pinned to different sockets), you can set I_MPI_PIN_ORDER=scatter.&lt;/P&gt;

&lt;P&gt;Our online Intel MPI Library Reference Manual has an entire section on the pinning schema and our OpenMP interoperability: &lt;A href="https://software.intel.com/en-us/node/528819" target="_blank"&gt;click here for link&lt;/A&gt;.&amp;nbsp; You're welcome to take a look.&lt;/P&gt;

&lt;P&gt;Regards,&lt;BR /&gt;
	~Gergana&lt;/P&gt;</description>
    <pubDate>Tue, 06 Jan 2015 22:46:23 GMT</pubDate>
    <dc:creator>Gergana_S_Intel</dc:creator>
    <dc:date>2015-01-06T22:46:23Z</dc:date>
    <item>
      <title>Intel MPI process placement</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038276#M4261</link>
      <description>&lt;DIV class="field field-name-comment-body field-type-text-long field-label-hidden"&gt;
	&lt;DIV class="field-items"&gt;
		&lt;DIV class="field-item even"&gt;
			&lt;P&gt;Hello,&lt;/P&gt;

			&lt;P&gt;I would like to pin MPI processes across all CPU sockets. For example, I would like to run 10 MPI processes on a two socket machine with 5 MPI processes on each socket. Could you please send me the instructions on doing this?&lt;/P&gt;

			&lt;P&gt;Many thanks,&lt;/P&gt;
		&lt;/DIV&gt;
	&lt;/DIV&gt;
&lt;/DIV&gt;</description>
      <pubDate>Thu, 01 Jan 2015 23:02:35 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038276#M4261</guid>
      <dc:creator>Miah__Wadud</dc:creator>
      <dc:date>2015-01-01T23:02:35Z</dc:date>
    </item>
    <item>
      <title>Hi Wadud,</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038277#M4262</link>
      <description>&lt;P&gt;Hi Wadud,&lt;/P&gt;

&lt;P&gt;To pin MPI process across all sockets, you just need to set I_MPI_PIN_DOMAIN=socket.&amp;nbsp; You can do that before you run your application (export I_MPI_PIN_DOMAIN=socket) or on the mpirun command line by using the -genv option (mpirun -genv I_MPI_PIN_DOMAIN socket -n 10 ./exe).&lt;/P&gt;

&lt;P&gt;To double-check if the pinning is correct, just set I_MPI_DEBUG=4 and Intel MPI will print a pinning table at the beginning of the run.&lt;/P&gt;

&lt;P&gt;Let me know how this works.&lt;/P&gt;

&lt;P&gt;Regards,&lt;BR /&gt;
	~Gergana&lt;/P&gt;</description>
      <pubDate>Tue, 06 Jan 2015 22:03:42 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038277#M4262</guid>
      <dc:creator>Gergana_S_Intel</dc:creator>
      <dc:date>2015-01-06T22:03:42Z</dc:date>
    </item>
    <item>
      <title>Hi Gergana,</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038278#M4263</link>
      <description>&lt;P&gt;Hi Gergana,&lt;/P&gt;

&lt;P&gt;thanks for the reply. Does the configuration I_MPI_PIN_DOMAIN=socket do a scatter, much like how it is done in OpenMP?&lt;/P&gt;

&lt;P&gt;Thanks.&lt;/P&gt;</description>
      <pubDate>Tue, 06 Jan 2015 22:25:43 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038278#M4263</guid>
      <dc:creator>Miah__Wadud</dc:creator>
      <dc:date>2015-01-06T22:25:43Z</dc:date>
    </item>
    <item>
      <title>I believe the default value</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038279#M4264</link>
      <description>&lt;P&gt;I believe the default value is actually 'compact'.&amp;nbsp; That means the first 5 consecutive ranks will be pinned to the first socket, and the next set of 5 will be pinning to the second socket.&lt;/P&gt;

&lt;P&gt;If you want to do 'scatter' instead (which, in your case, means alternating ranks will be pinned to different sockets), you can set I_MPI_PIN_ORDER=scatter.&lt;/P&gt;

&lt;P&gt;Our online Intel MPI Library Reference Manual has an entire section on the pinning schema and our OpenMP interoperability: &lt;A href="https://software.intel.com/en-us/node/528819" target="_blank"&gt;click here for link&lt;/A&gt;.&amp;nbsp; You're welcome to take a look.&lt;/P&gt;

&lt;P&gt;Regards,&lt;BR /&gt;
	~Gergana&lt;/P&gt;</description>
      <pubDate>Tue, 06 Jan 2015 22:46:23 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038279#M4264</guid>
      <dc:creator>Gergana_S_Intel</dc:creator>
      <dc:date>2015-01-06T22:46:23Z</dc:date>
    </item>
    <item>
      <title>Hi ,</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038280#M4265</link>
      <description>&lt;DIV class="field field-name-comment-body field-type-text-long field-label-hidden"&gt;
	&lt;DIV class="field-items"&gt;
		&lt;DIV class="field-item even"&gt;
			&lt;P&gt;Hi ,&lt;/P&gt;

			&lt;P&gt;I have compiled espresso with intel mpi and MKL library but&amp;nbsp; getting error Failure during collective error when ever it is working fine with openmpi.&lt;/P&gt;

			&lt;P&gt;is there problem with intel mpi&lt;/P&gt;

			&lt;P&gt;Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x516f460, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x5300310, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x6b295c0, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x67183d0, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x4f794c0, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				[0:n125] unexpected disconnect completion event from [22:n122]&lt;BR /&gt;
				Assertion failed in file ../../dapl_conn_rc.c at line 1128: 0&lt;BR /&gt;
				internal ABORT - process 0&lt;BR /&gt;
				Fatal error in PMPI_Bcast: Other MPI error, error stack:&lt;BR /&gt;
				PMPI_Bcast(2112)........: MPI_Bcast(buf=0x56bfe30, count=96, MPI_DOUBLE_PRECISION, root=4, comm=0x84000004) failed&lt;BR /&gt;
				MPIR_Bcast_impl(1670)...:&lt;BR /&gt;
				I_MPIR_Bcast_intra(1887): Failure during collective&lt;BR /&gt;
				MPIR_Bcast_intra(1524)..: Failure during collective&lt;BR /&gt;
				/var/spool/PBS/mom_priv/epilogue: line 30: kill: (5089) - No such process&lt;/P&gt;

			&lt;P&gt;Kindly help us for resolving this&lt;/P&gt;

			&lt;P&gt;Thanks&lt;BR /&gt;
				sanjiv&lt;/P&gt;
		&lt;/DIV&gt;
	&lt;/DIV&gt;
&lt;/DIV&gt;</description>
      <pubDate>Thu, 04 Jun 2015 10:07:26 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/Intel-MPI-process-placement/m-p/1038280#M4265</guid>
      <dc:creator>Sanjiv_T_</dc:creator>
      <dc:date>2015-06-04T10:07:26Z</dc:date>
    </item>
  </channel>
</rss>

