<?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 SVD memory spike in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796832#M2737</link>
    <description>Excellent, I'm very happy to know about this tool. The problem was elsewhere in my code. Thank you for pointing me to this function.</description>
    <pubDate>Mon, 25 Jun 2012 18:44:49 GMT</pubDate>
    <dc:creator>tinastone</dc:creator>
    <dc:date>2012-06-25T18:44:49Z</dc:date>
    <item>
      <title>SVD memory spike</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796830#M2735</link>
      <description>Hi there,&lt;BR /&gt;&lt;BR /&gt;I am using MKL for it SVD functionality in a C++ program. I notice a huge memory spike towards the end of computing the SVD. This is sometimes causing crashes, slowdowns, etc, when swapping occurs.&lt;BR /&gt;&lt;BR /&gt;my code call is the following:&lt;BR /&gt;
info = LAPACKE_sgesvd(LAPACK_ROW_MAJOR, 'S', 'S', m, n, src, n, s, u, m, vt, n, superb);&lt;BR /&gt;
&lt;BR /&gt;
My computer is a Dell Precision T7500 with 2 Xeon E5645 CPUs. Hyperthreading is enabled. I have 12 GB system RAM.&lt;BR /&gt;&lt;BR /&gt;My testing dataset is 150x190x1340 (real, single precision), which I "flatten" into 28500x1340. This uses very little memory (perhaps some few hundreds of MB) for most of the computing time for the SVD. However, there is a brief spike to more than 6 GB usage. What is causing this huge spike? Is there some way I can avoid it? It is limiting the data set size that I can use.&lt;BR /&gt;&lt;BR /&gt;</description>
      <pubDate>Fri, 22 Jun 2012 17:22:10 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796830#M2735</guid>
      <dc:creator>tinastone</dc:creator>
      <dc:date>2012-06-22T17:22:10Z</dc:date>
    </item>
    <item>
      <title>SVD memory spike</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796831#M2736</link>
      <description>did you check whatmkl_mem_stat( buf) returns in this case?</description>
      <pubDate>Sat, 23 Jun 2012 07:49:30 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796831#M2736</guid>
      <dc:creator>Gennady_F_Intel</dc:creator>
      <dc:date>2012-06-23T07:49:30Z</dc:date>
    </item>
    <item>
      <title>SVD memory spike</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796832#M2737</link>
      <description>Excellent, I'm very happy to know about this tool. The problem was elsewhere in my code. Thank you for pointing me to this function.</description>
      <pubDate>Mon, 25 Jun 2012 18:44:49 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/SVD-memory-spike/m-p/796832#M2737</guid>
      <dc:creator>tinastone</dc:creator>
      <dc:date>2012-06-25T18:44:49Z</dc:date>
    </item>
  </channel>
</rss>

