<?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: Speed of IPP in Intel® Integrated Performance Primitives</title>
    <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936045#M17177</link>
    <description>&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;P&gt;Hao, we keep improving IPP in both functionality and performance, so every next version should be better than previous one. At least it is expected, otherwise what the reason to make new version?&lt;/P&gt;
&lt;P&gt;Remember we deliver performance data (look at IPP	oolsperfsysdata folder)for almost all IPP functions within IPP package, so you can just take the latest version and compare data for functions you are interested in.&lt;/P&gt;
&lt;P&gt;Regards,&lt;BR /&gt; Vladimir&lt;/P&gt;&lt;P&gt;Message Edited by vdudnik on &lt;SPAN class="date_text"&gt;04-11-2006&lt;/SPAN&gt; &lt;SPAN class="time_text"&gt;10:45 PM&lt;/SPAN&gt;&lt;/P&gt;</description>
    <pubDate>Wed, 12 Apr 2006 01:28:53 GMT</pubDate>
    <dc:creator>Vladimir_Dudnik</dc:creator>
    <dc:date>2006-04-12T01:28:53Z</dc:date>
    <item>
      <title>Speed of IPP</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936042#M17174</link>
      <description>&lt;DIV&gt;&lt;/DIV&gt;&lt;DIV&gt;Hi,&lt;/DIV&gt;&lt;DIV&gt;I'm not sure if anyone has compared IPP with mmx before. I tested performance of IPP (v4.1)in Intel PM2.13GHz 1MRAM by SAD functions (4x4, 8x8, 16x16). I found it is a little bit slower than the mmx code. I feel kind of strange because I think IPP should be better than mmx/sse code.Is it because it checks theCPU type in each function?&lt;/DIV&gt;&lt;DIV&gt;Also,I'm very curious toknow if IPP functions are written by assembly like mmx/sse or by other means such as machine code.&lt;/DIV&gt;&lt;DIV&gt;Expecting and thank you very much for your answer!&lt;/DIV&gt;&lt;DIV&gt;Hao&lt;/DIV&gt;&lt;P&gt;Message Edited by hzhang08@gmail.com on &lt;SPAN class="date_text"&gt;04-07-2006&lt;/SPAN&gt;&lt;SPAN class="time_text"&gt;02:32 PM&lt;/SPAN&gt;&lt;/P&gt;</description>
      <pubDate>Sat, 08 Apr 2006 04:16:33 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936042#M17174</guid>
      <dc:creator>hzhang08</dc:creator>
      <dc:date>2006-04-08T04:16:33Z</dc:date>
    </item>
    <item>
      <title>Re: Speed of IPP</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936043#M17175</link>
      <description>&lt;DIV&gt;Hello,&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;Yes, IPP functions mostly optimized by hand with ASM code and knowledge of Intel micro-architecture.&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;Did you test with IPP v5.0 or the latest version, IPP v5.1?&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;Regards,&lt;BR /&gt; Vladimir&lt;/DIV&gt;</description>
      <pubDate>Wed, 12 Apr 2006 00:57:50 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936043#M17175</guid>
      <dc:creator>Vladimir_Dudnik</dc:creator>
      <dc:date>2006-04-12T00:57:50Z</dc:date>
    </item>
    <item>
      <title>Re: Speed of IPP</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936044#M17176</link>
      <description>&lt;DIV&gt;Vladimir:&lt;/DIV&gt;&lt;DIV&gt;I only tested IPP4.1.Are there many differences in terms of performance between these two versions?&lt;/DIV&gt;&lt;DIV&gt;Hao&lt;/DIV&gt;</description>
      <pubDate>Wed, 12 Apr 2006 01:19:19 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936044#M17176</guid>
      <dc:creator>hzhang08</dc:creator>
      <dc:date>2006-04-12T01:19:19Z</dc:date>
    </item>
    <item>
      <title>Re: Speed of IPP</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936045#M17177</link>
      <description>&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;P&gt;Hao, we keep improving IPP in both functionality and performance, so every next version should be better than previous one. At least it is expected, otherwise what the reason to make new version?&lt;/P&gt;
&lt;P&gt;Remember we deliver performance data (look at IPP	oolsperfsysdata folder)for almost all IPP functions within IPP package, so you can just take the latest version and compare data for functions you are interested in.&lt;/P&gt;
&lt;P&gt;Regards,&lt;BR /&gt; Vladimir&lt;/P&gt;&lt;P&gt;Message Edited by vdudnik on &lt;SPAN class="date_text"&gt;04-11-2006&lt;/SPAN&gt; &lt;SPAN class="time_text"&gt;10:45 PM&lt;/SPAN&gt;&lt;/P&gt;</description>
      <pubDate>Wed, 12 Apr 2006 01:28:53 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936045#M17177</guid>
      <dc:creator>Vladimir_Dudnik</dc:creator>
      <dc:date>2006-04-12T01:28:53Z</dc:date>
    </item>
    <item>
      <title>Re: Speed of IPP</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936046#M17178</link>
      <description>&lt;DIV&gt;This isvery helpful. I'm going to check it. Thank you Vladimir.&lt;/DIV&gt;</description>
      <pubDate>Wed, 12 Apr 2006 01:51:13 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/Speed-of-IPP/m-p/936046#M17178</guid>
      <dc:creator>hzhang08</dc:creator>
      <dc:date>2006-04-12T01:51:13Z</dc:date>
    </item>
  </channel>
</rss>

