<?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 spin lock in Intel mpi in Intel® MPI Library</title>
    <link>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916132#M2322</link>
    <description>I notice that when intel mpi is idle it is taking up 100% of one of my cpu's. Since my application is more computational intensive over multi-core's than communication intensive, this is causing me large performance problems. Is there any way I can get rid of this spin lock or control it's polling rate, even at the loss of latency?&lt;BR /&gt;&lt;BR /&gt;</description>
    <pubDate>Sun, 15 Jul 2007 05:23:27 GMT</pubDate>
    <dc:creator>David_M_13</dc:creator>
    <dc:date>2007-07-15T05:23:27Z</dc:date>
    <item>
      <title>spin lock in Intel mpi</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916132#M2322</link>
      <description>I notice that when intel mpi is idle it is taking up 100% of one of my cpu's. Since my application is more computational intensive over multi-core's than communication intensive, this is causing me large performance problems. Is there any way I can get rid of this spin lock or control it's polling rate, even at the loss of latency?&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Sun, 15 Jul 2007 05:23:27 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916132#M2322</guid>
      <dc:creator>David_M_13</dc:creator>
      <dc:date>2007-07-15T05:23:27Z</dc:date>
    </item>
    <item>
      <title>Re: spin lock in Intel mpi</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916133#M2323</link>
      <description>&lt;P&gt;Hi David,&lt;/P&gt;
&lt;P&gt;The Intel MPI Library provides a way to control the value of the spin count when polling the fabrics on a cluster by using the I_MPI_SPIN_COUNT env variable. Its argument should be the number of times the polling loop will spin before freeing the processors. Generally, a smaller value with release the processor more frequently.&lt;/P&gt;
&lt;P&gt;The default settings are looping 1 time for the sockets/shared memory devices and 250 times when using RDMA-based devices. You can try changing the value and see if it makes a difference for your application. More information is in the Intel MPI Library Reference Manual in the &lt;INSTALL_DIR&gt;/doc directory.&lt;/INSTALL_DIR&gt;&lt;/P&gt;
&lt;P&gt;Thanks,&lt;BR /&gt;~Gergana&lt;/P&gt;</description>
      <pubDate>Mon, 16 Jul 2007 14:44:47 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916133#M2323</guid>
      <dc:creator>Gergana_S_Intel</dc:creator>
      <dc:date>2007-07-16T14:44:47Z</dc:date>
    </item>
    <item>
      <title>Re: spin lock in Intel mpi</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916134#M2324</link>
      <description>Hi Gergana,&lt;BR /&gt;I played around with I_MPI_SPIN_COUNT already and it doesn't seem to make a difference. I've also run my application under MPICH2 and MVAPICH2 without this problem. &lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;David&lt;BR /&gt;</description>
      <pubDate>Tue, 17 Jul 2007 05:55:49 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916134#M2324</guid>
      <dc:creator>David_M_13</dc:creator>
      <dc:date>2007-07-17T05:55:49Z</dc:date>
    </item>
    <item>
      <title>Re: spin lock in Intel mpi</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916135#M2325</link>
      <description>&lt;P&gt;Hi David,&lt;/P&gt;
&lt;P&gt;Please try the latest Intel MPI Library 3.1 with the I_MPI_WAIT_MODE environment variable set to enable. It shouldwork for following the sock, shm, or ssmdevices. Let us know how the suggestion help.&lt;/P&gt;
&lt;P&gt;Best regards, Andrey&lt;/P&gt;</description>
      <pubDate>Mon, 12 Nov 2007 12:01:02 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/spin-lock-in-Intel-mpi/m-p/916135#M2325</guid>
      <dc:creator>Andrey_D_Intel</dc:creator>
      <dc:date>2007-11-12T12:01:02Z</dc:date>
    </item>
  </channel>
</rss>

