<?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 VML complex functions timing in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-complex-functions-timing/m-p/811321#M3957</link>
    <description>&lt;DIV&gt;Hi,&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;I was browsing the following page which lists the Clocks Per Element timing for all the VML functions.&lt;/DIV&gt;&lt;A href="http://software.intel.com/sites/products/documentation/hpc/mkl/vml3/functions/_performanceall.html"&gt;http://software.intel.com/sites/products/documentation/hpc/mkl/vml3/functions/_performanceall.html&lt;/A&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Does anyone else think the timings for the Complex functions seem suspect?&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;E.g.&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Real Add : 0.79&lt;/DIV&gt;&lt;DIV&gt;Complex Add : 19.11&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Surely Complex Add should only be 2x slower?&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;</description>
    <pubDate>Tue, 01 Feb 2011 06:58:30 GMT</pubDate>
    <dc:creator>Shueng_Chuan_Kiu</dc:creator>
    <dc:date>2011-02-01T06:58:30Z</dc:date>
    <item>
      <title>VML complex functions timing</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-complex-functions-timing/m-p/811321#M3957</link>
      <description>&lt;DIV&gt;Hi,&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;I was browsing the following page which lists the Clocks Per Element timing for all the VML functions.&lt;/DIV&gt;&lt;A href="http://software.intel.com/sites/products/documentation/hpc/mkl/vml3/functions/_performanceall.html"&gt;http://software.intel.com/sites/products/documentation/hpc/mkl/vml3/functions/_performanceall.html&lt;/A&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Does anyone else think the timings for the Complex functions seem suspect?&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;E.g.&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Real Add : 0.79&lt;/DIV&gt;&lt;DIV&gt;Complex Add : 19.11&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Surely Complex Add should only be 2x slower?&lt;/DIV&gt;&lt;DIV&gt;&lt;/DIV&gt;</description>
      <pubDate>Tue, 01 Feb 2011 06:58:30 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-complex-functions-timing/m-p/811321#M3957</guid>
      <dc:creator>Shueng_Chuan_Kiu</dc:creator>
      <dc:date>2011-02-01T06:58:30Z</dc:date>
    </item>
    <item>
      <title>VML complex functions timing</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-complex-functions-timing/m-p/811322#M3958</link>
      <description>Complex add/sub functions support FP overflow error handling, while real add/sub functions do not guarantee that. This is the reason for difference in performance.</description>
      <pubDate>Tue, 01 Feb 2011 11:42:35 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-complex-functions-timing/m-p/811322#M3958</guid>
      <dc:creator>Ilya_B_Intel</dc:creator>
      <dc:date>2011-02-01T11:42:35Z</dc:date>
    </item>
  </channel>
</rss>

