<?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 Re: VML performance in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975977#M17095</link>
    <description>This presentation is strange, that's for sure.  Hard to believe the equivalent measures are being compared.  &lt;BR /&gt;&lt;BR /&gt;As the comparisons between P-III and P4 double libraries confirm, you would expect more clocks to be required when x87 extended precision is used to obtain a fully accurate result.  Also confirmed is an increased number of clocks for the latest style of P4, compared with the earlier one.&lt;BR /&gt;&lt;BR /&gt;I assume the distinction between the 2 columns is according to whether the 32- or 64-bit library is in use, so you could take your choice on Pentium-D as well.  I suppose you could run all versions of the library on your Pentium-D with 64-bit OS, and check the tables yourself.</description>
    <pubDate>Thu, 15 Sep 2005 03:34:36 GMT</pubDate>
    <dc:creator>TimP</dc:creator>
    <dc:date>2005-09-15T03:34:36Z</dc:date>
    <item>
      <title>VML performance</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975976#M17094</link>
      <description>Hi,&lt;BR /&gt;&lt;BR /&gt;I just downloaded MKL. Browsing the online documentation, I see that the performance of some VML functions (atan2 for instance) is significantly slower (even a factor of 10, or 20, or worse) on Intel Xeon EM64T than on other architectures (see &lt;A href="https://community.intel.com/www.intel.com/software/products/mkl/data/vml/functions/atan2.html)" target="_blank"&gt;www.intel.com/software/products/mkl/data/vml/functions/atan2.html)&lt;/A&gt;.&lt;BR /&gt;&lt;BR /&gt;- Does that simply mean that some of the VML functions have not been optimized for that architecture yet?&lt;BR /&gt;- What would a Pentium D be on that performance table? A P4 with SSE3 or a Xeon with EM64T?&lt;BR /&gt;&lt;BR /&gt;Thanks!</description>
      <pubDate>Wed, 14 Sep 2005 06:32:00 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975976#M17094</guid>
      <dc:creator>olecram</dc:creator>
      <dc:date>2005-09-14T06:32:00Z</dc:date>
    </item>
    <item>
      <title>Re: VML performance</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975977#M17095</link>
      <description>This presentation is strange, that's for sure.  Hard to believe the equivalent measures are being compared.  &lt;BR /&gt;&lt;BR /&gt;As the comparisons between P-III and P4 double libraries confirm, you would expect more clocks to be required when x87 extended precision is used to obtain a fully accurate result.  Also confirmed is an increased number of clocks for the latest style of P4, compared with the earlier one.&lt;BR /&gt;&lt;BR /&gt;I assume the distinction between the 2 columns is according to whether the 32- or 64-bit library is in use, so you could take your choice on Pentium-D as well.  I suppose you could run all versions of the library on your Pentium-D with 64-bit OS, and check the tables yourself.</description>
      <pubDate>Thu, 15 Sep 2005 03:34:36 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975977#M17095</guid>
      <dc:creator>TimP</dc:creator>
      <dc:date>2005-09-15T03:34:36Z</dc:date>
    </item>
    <item>
      <title>Re: VML performance</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975978#M17096</link>
      <description>&lt;DIV&gt;
&lt;DIV&gt;&lt;FONT size="2"&gt;VML part of MKL 7.2 has several functions, which are not optimized for EM64T. &lt;BR /&gt;The reason is that initially Microsoft announced to drop x87 support in the OS. &lt;BR /&gt;So we were forced to quickly get rid of x87-based functions by temporarily replacing&lt;BR /&gt;them with rather slow portable implementations. &lt;BR /&gt;Improvements for some of these functions will be in MKL 8.0 &lt;BR /&gt;(e.g. for Tan, Atan2, Asinh, Acosh for certain floating-point format and accuracy flavor) &lt;BR /&gt;while for other remaining functions it will be done in subsequent releases.&lt;/FONT&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;FONT size="2"&gt;&lt;/FONT&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;FONT size="2"&gt;As to your second question about VML performance data for Pentium D processor, &lt;BR /&gt;we do not publish this data on the web yet. &lt;BR /&gt;Im not sure that existing performance columns will help you getting absolutely &lt;BR /&gt;adequate performance information on Pentium D processor.&lt;/FONT&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;FONT size="2"&gt;Thanks,&lt;BR /&gt;Andrey&lt;BR /&gt;&lt;/FONT&gt;&lt;/DIV&gt;&lt;/DIV&gt;</description>
      <pubDate>Thu, 15 Sep 2005 15:07:49 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/VML-performance/m-p/975978#M17096</guid>
      <dc:creator>Andrey_N_Intel</dc:creator>
      <dc:date>2005-09-15T15:07:49Z</dc:date>
    </item>
  </channel>
</rss>

