<?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 Ranks within a node can in Intel® MPI Library</title>
    <link>https://community.intel.com/t5/Intel-MPI-Library/How-communication-between-ranks-works/m-p/922589#M2402</link>
    <description>&lt;P&gt;Ranks within a node can exchange messages directly. &amp;nbsp;Usually, you would have enabled shm (a normal default) so that this is done memcpy style.&lt;/P&gt;</description>
    <pubDate>Wed, 05 Mar 2014 14:54:00 GMT</pubDate>
    <dc:creator>TimP</dc:creator>
    <dc:date>2014-03-05T14:54:00Z</dc:date>
    <item>
      <title>How communication between ranks works?</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/How-communication-between-ranks-works/m-p/922588#M2401</link>
      <description>&lt;P&gt;Example of hipo cluster :&amp;nbsp;&lt;/P&gt;

&lt;P&gt;"Node 1" -&amp;gt; 4 cores so [0,4[ rank&lt;/P&gt;

&lt;P&gt;"Node 2" -&amp;gt; 4 cores so [4,8[ rank&lt;/P&gt;

&lt;P&gt;So rank = 0 would be first core in Node 1&lt;/P&gt;

&lt;P&gt;In a situation like Core #5 wants to send a message to Core #6&lt;/P&gt;

&lt;P&gt;which are both in "Node 2"&lt;/P&gt;

&lt;P&gt;What happens?&lt;/P&gt;

&lt;P&gt;&lt;SPAN style="font-size: 1em; line-height: 1.5;"&gt;I hope Core 5 doesn't sends its message to Rank #0 in "Node 1" that sends it back to Core #6.&lt;/SPAN&gt;&lt;/P&gt;

&lt;P&gt;So how does it works?&lt;/P&gt;</description>
      <pubDate>Sat, 22 Feb 2014 18:36:25 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/How-communication-between-ranks-works/m-p/922588#M2401</guid>
      <dc:creator>Carlos_Alexandre_T_</dc:creator>
      <dc:date>2014-02-22T18:36:25Z</dc:date>
    </item>
    <item>
      <title>Ranks within a node can</title>
      <link>https://community.intel.com/t5/Intel-MPI-Library/How-communication-between-ranks-works/m-p/922589#M2402</link>
      <description>&lt;P&gt;Ranks within a node can exchange messages directly. &amp;nbsp;Usually, you would have enabled shm (a normal default) so that this is done memcpy style.&lt;/P&gt;</description>
      <pubDate>Wed, 05 Mar 2014 14:54:00 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-MPI-Library/How-communication-between-ranks-works/m-p/922589#M2402</guid>
      <dc:creator>TimP</dc:creator>
      <dc:date>2014-03-05T14:54:00Z</dc:date>
    </item>
  </channel>
</rss>

