<?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 Do PMUs occasionally lose their reliability? in Software Tuning, Performance Optimization &amp; Platform Monitoring</title>
    <link>https://community.intel.com/t5/Software-Tuning-Performance/Do-PMUs-occasionally-lose-their-reliability/m-p/1128850#M6367</link>
    <description>&lt;P&gt;Hello all,&lt;/P&gt;

&lt;P&gt;I ran a few tests with VTune&amp;nbsp;and I found results&amp;nbsp;not as they're supposed to be.&lt;/P&gt;

&lt;P&gt;lbm_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 362,259,200,000&lt;BR /&gt;
	lbm_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,261,966,300,000&lt;BR /&gt;
	lbm_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,261,717,900,000&lt;BR /&gt;
	lbm_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,262,373,400,000&lt;/P&gt;

&lt;P&gt;leslie3d_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,202,847,100,000&lt;BR /&gt;
	leslie3d_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 824,536,200,000&lt;BR /&gt;
	leslie3d_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,552,713,900,000&lt;BR /&gt;
	leslie3d_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,553,852,400,000&lt;/P&gt;

&lt;P&gt;libquantum_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 2,307,256,500,000&lt;BR /&gt;
	libquantum_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 484,984,900,000&lt;BR /&gt;
	libquantum_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,124,465,400,000&lt;BR /&gt;
	libquantum_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 2,307,302,500,000&lt;/P&gt;

&lt;P&gt;Here are examples of data that went wrong. Son/off refers to SMT(Hyperthreading) on/off. 2p1c refers to 2 threads running on 1 core, and similarly 2p2c refers to 2 threads running on&amp;nbsp;separated cores. These are simple setups&amp;nbsp;of my experiment. My experiment platform is intel E5 2650 v3(Haswell), CentOS 7.0.&lt;/P&gt;

&lt;P&gt;Regardless the running environment, the total instructions retired by one single completely run process&amp;nbsp;is supposed to be permanent(is this right?). Since Vtune gave a MUX reliability higher than 0.995, I suppose such mis-counts are due to the PMUs themselves. Is there any chance PMUs suffer from external impacts such as raised temperature and cause counting failures? Do other datas given by PMUs have similar risk to be trusted?&lt;/P&gt;

&lt;P&gt;Thanks!&lt;/P&gt;</description>
    <pubDate>Fri, 22 Sep 2017 03:24:16 GMT</pubDate>
    <dc:creator>CciYu_C_</dc:creator>
    <dc:date>2017-09-22T03:24:16Z</dc:date>
    <item>
      <title>Do PMUs occasionally lose their reliability?</title>
      <link>https://community.intel.com/t5/Software-Tuning-Performance/Do-PMUs-occasionally-lose-their-reliability/m-p/1128850#M6367</link>
      <description>&lt;P&gt;Hello all,&lt;/P&gt;

&lt;P&gt;I ran a few tests with VTune&amp;nbsp;and I found results&amp;nbsp;not as they're supposed to be.&lt;/P&gt;

&lt;P&gt;lbm_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 362,259,200,000&lt;BR /&gt;
	lbm_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,261,966,300,000&lt;BR /&gt;
	lbm_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,261,717,900,000&lt;BR /&gt;
	lbm_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,262,373,400,000&lt;/P&gt;

&lt;P&gt;leslie3d_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,202,847,100,000&lt;BR /&gt;
	leslie3d_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 824,536,200,000&lt;BR /&gt;
	leslie3d_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,552,713,900,000&lt;BR /&gt;
	leslie3d_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,553,852,400,000&lt;/P&gt;

&lt;P&gt;libquantum_Soff_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 2,307,256,500,000&lt;BR /&gt;
	libquantum_Soff_2p2c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 484,984,900,000&lt;BR /&gt;
	libquantum_Son_1p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 1,124,465,400,000&lt;BR /&gt;
	libquantum_Son_2p1c:&amp;nbsp;&amp;nbsp;&amp;nbsp; Instructions Retired: 2,307,302,500,000&lt;/P&gt;

&lt;P&gt;Here are examples of data that went wrong. Son/off refers to SMT(Hyperthreading) on/off. 2p1c refers to 2 threads running on 1 core, and similarly 2p2c refers to 2 threads running on&amp;nbsp;separated cores. These are simple setups&amp;nbsp;of my experiment. My experiment platform is intel E5 2650 v3(Haswell), CentOS 7.0.&lt;/P&gt;

&lt;P&gt;Regardless the running environment, the total instructions retired by one single completely run process&amp;nbsp;is supposed to be permanent(is this right?). Since Vtune gave a MUX reliability higher than 0.995, I suppose such mis-counts are due to the PMUs themselves. Is there any chance PMUs suffer from external impacts such as raised temperature and cause counting failures? Do other datas given by PMUs have similar risk to be trusted?&lt;/P&gt;

&lt;P&gt;Thanks!&lt;/P&gt;</description>
      <pubDate>Fri, 22 Sep 2017 03:24:16 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Tuning-Performance/Do-PMUs-occasionally-lose-their-reliability/m-p/1128850#M6367</guid>
      <dc:creator>CciYu_C_</dc:creator>
      <dc:date>2017-09-22T03:24:16Z</dc:date>
    </item>
    <item>
      <title>You may want to double-check</title>
      <link>https://community.intel.com/t5/Software-Tuning-Performance/Do-PMUs-occasionally-lose-their-reliability/m-p/1128851#M6368</link>
      <description>&lt;P&gt;You may want to double-check that the NMI watchdog is disabled.&amp;nbsp; VTune should do this, but....&lt;/P&gt;

&lt;P&gt;CentOS 7.0 is pretty old -- can you upgrade to a newer version?&amp;nbsp; (I found some bugs in Intel contributions to the kernel code in 7.2 that are fixed (sort of) in 7.3, and other bugs in 7.3 that only seem to apply to the newer Xeon Scalable Processors.&amp;nbsp; I have not tested CentOS 7.4.)&lt;/P&gt;</description>
      <pubDate>Fri, 22 Sep 2017 14:50:23 GMT</pubDate>
      <guid>https://community.intel.com/t5/Software-Tuning-Performance/Do-PMUs-occasionally-lose-their-reliability/m-p/1128851#M6368</guid>
      <dc:creator>McCalpinJohn</dc:creator>
      <dc:date>2017-09-22T14:50:23Z</dc:date>
    </item>
  </channel>
</rss>

