<?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: FFT2 has double version? in Intel® Integrated Performance Primitives</title>
    <link>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976149#M21012</link>
    <description>&lt;DIV&gt;Hi, &lt;/DIV&gt;
&lt;DIV&gt;Thank you very much for your response.&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;I am doing something on image processing. In this project, the image size is 4096X4096 or above, I have to perform the FFTs of 4-6 images andpeferform some operations among them, so, sometimes, the result became overflow. If in double type, no overflow,&lt;EM&gt; &lt;/EM&gt;I think.&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;(I have tried to divide a constant after FFT, but it is possible to lose accuracy, is not it?).&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;Best reagrads,&lt;/DIV&gt;
&lt;DIV&gt;Li&lt;/DIV&gt;</description>
    <pubDate>Thu, 08 Jan 2004 14:15:39 GMT</pubDate>
    <dc:creator>junli</dc:creator>
    <dc:date>2004-01-08T14:15:39Z</dc:date>
    <item>
      <title>FFT2 has double version?</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976147#M21010</link>
      <description>&lt;DIV&gt;Hi, &lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;In IPP 3.0, FFT2 only suports float data format. double is supoorte in IPP 4.0?&lt;/DIV&gt;</description>
      <pubDate>Tue, 06 Jan 2004 15:25:12 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976147#M21010</guid>
      <dc:creator>junli</dc:creator>
      <dc:date>2004-01-06T15:25:12Z</dc:date>
    </item>
    <item>
      <title>Re: FFT2 has double version?</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976148#M21011</link>
      <description>Dear Customer,&lt;BR /&gt;
&lt;DIV&gt;Double type is not supported in Intel IPP 4.0 neither, what is your concern for this data type in 2D FFT?&lt;/DIV&gt;
&lt;DIV&gt;&lt;BR /&gt;Thanks,&lt;BR /&gt;Ying &lt;/DIV&gt;</description>
      <pubDate>Thu, 08 Jan 2004 13:27:03 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976148#M21011</guid>
      <dc:creator>Ying_S_Intel</dc:creator>
      <dc:date>2004-01-08T13:27:03Z</dc:date>
    </item>
    <item>
      <title>Re: FFT2 has double version?</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976149#M21012</link>
      <description>&lt;DIV&gt;Hi, &lt;/DIV&gt;
&lt;DIV&gt;Thank you very much for your response.&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;I am doing something on image processing. In this project, the image size is 4096X4096 or above, I have to perform the FFTs of 4-6 images andpeferform some operations among them, so, sometimes, the result became overflow. If in double type, no overflow,&lt;EM&gt; &lt;/EM&gt;I think.&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;(I have tried to divide a constant after FFT, but it is possible to lose accuracy, is not it?).&lt;/DIV&gt;
&lt;DIV&gt;&lt;/DIV&gt;
&lt;DIV&gt;Best reagrads,&lt;/DIV&gt;
&lt;DIV&gt;Li&lt;/DIV&gt;</description>
      <pubDate>Thu, 08 Jan 2004 14:15:39 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976149#M21012</guid>
      <dc:creator>junli</dc:creator>
      <dc:date>2004-01-08T14:15:39Z</dc:date>
    </item>
    <item>
      <title>Re: FFT2 has double version?</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976150#M21013</link>
      <description>&lt;DIV&gt;I'm afraid you will get mem overflow having several images 4Kx4K of double&lt;/DIV&gt;</description>
      <pubDate>Sat, 31 Jan 2004 20:50:40 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/FFT2-has-double-version/m-p/976150#M21013</guid>
      <dc:creator>borix</dc:creator>
      <dc:date>2004-01-31T20:50:40Z</dc:date>
    </item>
  </channel>
</rss>

