<?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 Custom DLL - CPU specialization / performance in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Custom-DLL-CPU-specialization-performance/m-p/1633068#M36465</link>
    <description>&lt;P&gt;Dear support,&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;if I use "normal" MKL, some DLLs are loaded at runtime, depending on the CPU / instruction set - for instance, "mkl_avx2.2.dll" or "mkl_avx512.2.dll".&lt;/P&gt;&lt;P&gt;But what happens if I build a custom DLL? Will the makefile magically include all versions for all instruction sets (for each exported function), or will I fall back to some lowest common denominator?&lt;/P&gt;&lt;P&gt;In other words: Will the performance of a custom DLL match the performance of "normal" MKL?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks,&lt;/P&gt;&lt;P&gt;Jörn&lt;/P&gt;</description>
    <pubDate>Tue, 24 Sep 2024 08:57:38 GMT</pubDate>
    <dc:creator>Jörn</dc:creator>
    <dc:date>2024-09-24T08:57:38Z</dc:date>
    <item>
      <title>Custom DLL - CPU specialization / performance</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Custom-DLL-CPU-specialization-performance/m-p/1633068#M36465</link>
      <description>&lt;P&gt;Dear support,&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;if I use "normal" MKL, some DLLs are loaded at runtime, depending on the CPU / instruction set - for instance, "mkl_avx2.2.dll" or "mkl_avx512.2.dll".&lt;/P&gt;&lt;P&gt;But what happens if I build a custom DLL? Will the makefile magically include all versions for all instruction sets (for each exported function), or will I fall back to some lowest common denominator?&lt;/P&gt;&lt;P&gt;In other words: Will the performance of a custom DLL match the performance of "normal" MKL?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks,&lt;/P&gt;&lt;P&gt;Jörn&lt;/P&gt;</description>
      <pubDate>Tue, 24 Sep 2024 08:57:38 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Custom-DLL-CPU-specialization-performance/m-p/1633068#M36465</guid>
      <dc:creator>Jörn</dc:creator>
      <dc:date>2024-09-24T08:57:38Z</dc:date>
    </item>
  </channel>
</rss>

