<?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: IPP Performance Slow Down 10 Times in Intel® Integrated Performance Primitives</title>
    <link>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895530#M12019</link>
    <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
&lt;BR /&gt;If you're statically linking with IPP be sure to call ippStaticInit() otherwise the least optimized version of IPP routines will be used.&lt;BR /&gt;</description>
    <pubDate>Fri, 14 Aug 2009 15:39:31 GMT</pubDate>
    <dc:creator>pvonkaenel</dc:creator>
    <dc:date>2009-08-14T15:39:31Z</dc:date>
    <item>
      <title>IPP Performance Slow Down 10 Times</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895527#M12016</link>
      <description>I am using ippAffine to align images of 5M pixel 12bit mono image. The computer is Dell Precision T3400 which use E8400 CPU Core2 Duo. The normal speed is 10-20ms for one operation in testing code. However, its speed is 10 to 20 times slower(200-500ms) in our application software, in which two threads running two Affine on each of them simultaneously. Does anyone run into this kind of problem? Your insight is appreciated.&lt;BR /&gt;Yang</description>
      <pubDate>Thu, 13 Aug 2009 17:36:46 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895527#M12016</guid>
      <dc:creator>yangzhang000</dc:creator>
      <dc:date>2009-08-13T17:36:46Z</dc:date>
    </item>
    <item>
      <title>Re: IPP Performance Slow Down 10 Times</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895528#M12017</link>
      <description>&lt;DIV style="margin:0px;"&gt;
&lt;DIV id="quote_reply" style="margin-top: 5px; width: 100%;"&gt;
&lt;DIV style="margin-left:2px;margin-right:2px;"&gt;Quoting - &lt;A href="https://community.intel.com/en-us/profile/439009"&gt;yangzhang000&lt;/A&gt;&lt;/DIV&gt;
&lt;DIV style="background-color:#E5E5E5; padding:5px;border: 1px; border-style: inset;margin-left:2px;margin-right:2px;"&gt;&lt;EM&gt;I am using ippAffine to align images of 5M pixel 12bit mono image. The computer is Dell Precision T3400 which use E8400 CPU Core2 Duo. The normal speed is 10-20ms for one operation in testing code. However, its speed is 10 to 20 times slower(200-500ms) in our application software, in which two threads running two Affine on each of them simultaneously. Does anyone run into this kind of problem? Your insight is appreciated.&lt;BR /&gt;Yang&lt;/EM&gt;&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;BR /&gt;I doubted if it islibrary dispatcher issue. Which Intel IPP version do you use? and how do you link with Intel IPP libraires, if you link with static library, do you link threaded vesion or non-threaded version?&lt;BR /&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Ying S&lt;BR /&gt;</description>
      <pubDate>Thu, 13 Aug 2009 21:34:00 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895528#M12017</guid>
      <dc:creator>Ying_S_Intel</dc:creator>
      <dc:date>2009-08-13T21:34:00Z</dc:date>
    </item>
    <item>
      <title>Re: IPP Performance Slow Down 10 Times</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895529#M12018</link>
      <description>&lt;DIV style="margin:0px;"&gt;
&lt;DIV id="quote_reply" style="width: 100%; margin-top: 5px;"&gt;
&lt;DIV style="margin-left:2px;margin-right:2px;"&gt;Quoting - &lt;A href="https://community.intel.com/en-us/profile/336671"&gt;Ying S (Intel)&lt;/A&gt;&lt;/DIV&gt;
&lt;DIV style="background-color:#E5E5E5; padding:5px;border: 1px; border-style: inset;margin-left:2px;margin-right:2px;"&gt;&lt;EM&gt;
&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
&lt;BR /&gt;I doubted if it islibrary dispatcher issue. Which Intel IPP version do you use? and how do you link with Intel IPP libraires, if you link with static library, do you link threaded vesion or non-threaded version?&lt;BR /&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Ying S&lt;BR /&gt;&lt;/EM&gt;&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;BR /&gt;Hi,&lt;BR /&gt;If you're using a threaded version of ipp (either DLL or static threaded), you might try ippSetNumThreads(1)&lt;BR /&gt;You're already doing the multithreading from the application so if ipp is trying to do it also, you get more threads than CPU cores leading to much-much overhead. I ran in this once, 100% CPU usage returned to 30% on a live stream processing...&lt;BR /&gt;&lt;BR /&gt;Hope this helps.&lt;BR /&gt;&lt;BR /&gt;Matthieu&lt;BR /&gt;</description>
      <pubDate>Fri, 14 Aug 2009 07:42:56 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895529#M12018</guid>
      <dc:creator>matthieu_darbois</dc:creator>
      <dc:date>2009-08-14T07:42:56Z</dc:date>
    </item>
    <item>
      <title>Re: IPP Performance Slow Down 10 Times</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895530#M12019</link>
      <description>&lt;DIV style="margin:0px;"&gt;&lt;/DIV&gt;
&lt;BR /&gt;If you're statically linking with IPP be sure to call ippStaticInit() otherwise the least optimized version of IPP routines will be used.&lt;BR /&gt;</description>
      <pubDate>Fri, 14 Aug 2009 15:39:31 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/IPP-Performance-Slow-Down-10-Times/m-p/895530#M12019</guid>
      <dc:creator>pvonkaenel</dc:creator>
      <dc:date>2009-08-14T15:39:31Z</dc:date>
    </item>
  </channel>
</rss>

