<?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:pardiso residual in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/pardiso-residual/m-p/1578351#M35894</link>
    <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;Generally, if your input matrices with large condition numbers, then solver matrices would have better accuracy output. Can you check your input matrix? And could you please attach a simple reproducer, then you might get help from other MKL users in this community?&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;BR /&gt;</description>
    <pubDate>Thu, 07 Mar 2024 02:24:20 GMT</pubDate>
    <dc:creator>Ruqiu_C_Intel</dc:creator>
    <dc:date>2024-03-07T02:24:20Z</dc:date>
    <item>
      <title>pardiso residual</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/pardiso-residual/m-p/1575341#M35871</link>
      <description>&lt;P&gt;Dear mkl experts,&lt;/P&gt;&lt;P&gt;I get a problem when I am trying to solve a linear system with pardiso. The relative residual(||Ax-b||/||b||, A is a 2000000 * 2000000 sparse matrix)is always maintained at 0.001072 and I want to get a more accurate result. Here is the iparm values:&lt;/P&gt;&lt;P&gt;iparm[0] = 1;&amp;nbsp;&lt;BR /&gt;iparm[1] = 3;&lt;/P&gt;&lt;P&gt;iparm[7] = 100;&lt;/P&gt;&lt;P&gt;iparm[9] = 13;&lt;/P&gt;&lt;P&gt;iparm[10] = 1;&lt;/P&gt;&lt;P&gt;iparm[12] = 1;&lt;/P&gt;&lt;P&gt;iparm[17] = -1;&lt;/P&gt;&lt;P&gt;iparm[18] = -1;&lt;/P&gt;&lt;P&gt;iparm[23] = 1;&lt;/P&gt;&lt;P&gt;iparm[26&lt;SPAN&gt;] = 1;&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;iparm[31] = 1;&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;iparm[33] = 8;&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;iparm[34] = 1;&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN&gt;How should I adjust the iparm values? Your help &lt;SPAN&gt;appreciated.&lt;/SPAN&gt;&lt;/SPAN&gt;&lt;/P&gt;</description>
      <pubDate>Mon, 26 Feb 2024 08:57:14 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/pardiso-residual/m-p/1575341#M35871</guid>
      <dc:creator>Humphrey</dc:creator>
      <dc:date>2024-02-26T08:57:14Z</dc:date>
    </item>
    <item>
      <title>Re:pardiso residual</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/pardiso-residual/m-p/1578351#M35894</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;Generally, if your input matrices with large condition numbers, then solver matrices would have better accuracy output. Can you check your input matrix? And could you please attach a simple reproducer, then you might get help from other MKL users in this community?&lt;/P&gt;&lt;P&gt;&lt;BR /&gt;&lt;/P&gt;&lt;BR /&gt;</description>
      <pubDate>Thu, 07 Mar 2024 02:24:20 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/pardiso-residual/m-p/1578351#M35894</guid>
      <dc:creator>Ruqiu_C_Intel</dc:creator>
      <dc:date>2024-03-07T02:24:20Z</dc:date>
    </item>
  </channel>
</rss>

