<?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 How should CHA counters be counted？ in Software Tuning, Performance Optimization &amp; Platform Monitoring</title>
    <link>https://community.intel.com/t5/Software-Tuning-Performance/How-should-CHA-counters-be-counted/m-p/1654078#M8493</link>
    <description>&lt;P&gt;Hi，&lt;/P&gt;&lt;P&gt;&amp;nbsp; &amp;nbsp; The Intel Xeon Gold 6256 processor was tested, which has 12 physical cores. When the CHA counters was used to monitor the uncore performance event UNC_CHA_TOR_OCCUPANCY.IO_HIT/MISS, it found that CHA0-23 could collect data (not all counters have a value). CHA counters with an index greater than 24 always have a value of 0. This processor has the same number of CHA counters as the number of physical cores, so should I count only the CHA0-11 counters? Or counting the CHA0-23 counter?&lt;/P&gt;</description>
    <pubDate>Fri, 03 Jan 2025 12:39:28 GMT</pubDate>
    <dc:creator>Hz12</dc:creator>
    <dc:date>2025-01-03T12:39:28Z</dc:date>
    <item>
      <title>How should CHA counters be counted？</title>
      <link>https://community.intel.com/t5/Software-Tuning-Performance/How-should-CHA-counters-be-counted/m-p/1654078#M8493</link>
      <description>&lt;P&gt;Hi，&lt;/P&gt;&lt;P&gt;&amp;nbsp; &amp;nbsp; The Intel Xeon Gold 6256 processor was tested, which has 12 physical cores. When the CHA counters was used to monitor the uncore performance event UNC_CHA_TOR_OCCUPANCY.IO_HIT/MISS, it found that CHA0-23 could collect data (not all counters have a value). CHA counters with an index greater than 24 always have a value of 0. This processor has the same number of CHA counters as the number of physical cores, so should I count only the CHA0-11 counters? Or counting the CHA0-23 counter?&lt;/P&gt;</description>
      <pubDate>Fri, 03 Jan 2025 12:39:28 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Tuning-Performance/How-should-CHA-counters-be-counted/m-p/1654078#M8493</guid>
      <dc:creator>Hz12</dc:creator>
      <dc:date>2025-01-03T12:39:28Z</dc:date>
    </item>
    <item>
      <title>Re: How should CHA counters be counted？</title>
      <link>https://community.intel.com/t5/Software-Tuning-Performance/How-should-CHA-counters-be-counted/m-p/1660917#M8508</link>
      <description>&lt;P&gt;According to &lt;A href="https://www.intel.com/content/www/us/en/products/sku/198655/intel-xeon-gold-6256-processor-33m-cache-3-60-ghz/specifications.html" target="_blank"&gt;https://www.intel.com/content/www/us/en/products/sku/198655/intel-xeon-gold-6256-processor-33m-cache-3-60-ghz/specifications.html&lt;/A&gt;, the&amp;nbsp;Xeon Gold 6256 processor has 33 MiB of L3 cache. &amp;nbsp;For the SKX and CLX generations, each L3 "slice" is 1.375 MiB, so 33 MiB total implies that 24 L3 slices are active. &amp;nbsp;&lt;/P&gt;&lt;P&gt;So there is no problem -- measure and use results from CHAs 0 to 23.&lt;/P&gt;</description>
      <pubDate>Mon, 27 Jan 2025 16:12:08 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Tuning-Performance/How-should-CHA-counters-be-counted/m-p/1660917#M8508</guid>
      <dc:creator>McCalpinJohn</dc:creator>
      <dc:date>2025-01-27T16:12:08Z</dc:date>
    </item>
  </channel>
</rss>

