https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977855#M17194 <P>Hello</P> <P>I have a signal processing task which the heaviest operation is finding the eigenvalues and eigenvectors of a 128x128 Hermitian matrix. I found routines both in ipp and in mkl. Which one will be faster? I am using standard i7 windows 64 with 16Gb memory.&nbsp;</P> <P>Regards</P> <P>Dani</P> Sun, 18 Aug 2013 15:01:54 GMT Dani_Raphaeli 2013-08-18T15:01:54Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977855#M17194 <P>Hello</P> <P>I have a signal processing task which the heaviest operation is finding the eigenvalues and eigenvectors of a 128x128 Hermitian matrix. I found routines both in ipp and in mkl. Which one will be faster? I am using standard i7 windows 64 with 16Gb memory.&nbsp;</P> <P>Regards</P> <P>Dani</P> Sun, 18 Aug 2013 15:01:54 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977855#M17194 Dani_Raphaeli 2013-08-18T15:01:54Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977856#M17195 <P>Dani,</P> <P>MKL is a better choice, as IPP is typically optimized for very small matrix size (3x3, 4x4, etc). For such size of problem, you can choose MKL.</P> <P>Thanks,<BR />Chao</P> Wed, 21 Aug 2013 02:44:40 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977856#M17195 Chao_Y_Intel 2013-08-21T02:44:40Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977857#M17196 <STRONG>IPP</STRONG> has a set of functions for finding eigenvalues of symmetric and non-symmetric matricies and take a look at a <STRONG>ippm.h</STRONG> header file. Thu, 22 Aug 2013 05:10:21 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-vs-ipp-for-eigenvalues/m-p/977857#M17196 SergeyKostrov 2013-08-22T05:10:21Z