https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Is-mkl-dbsrmm-parallel/m-p/803106#M3224 Hello everyone,<DIV><SPAN style="font-family: Verdana, Arial, Helvetica, sans-serif;"><BR /></SPAN> <DIV>I was wondering whether <CODE>mkl_dbsrmm</CODE> has a parallel implementation. I have set both <CODE>OMP_NUM_THREADS</CODE> and <CODE>MKL_NUM_THREADS</CODE> to 8 and I checked that number with<PRE>[fortran] nthr = mkl_domain_get_max_threads( MKL_BLAS )[/fortran]</PRE> </DIV><DIV></DIV><DIV>However, when I check the CPU usage with top it never goes above 100%. I remember seeing higher CPU usage (400%) when using <CODE>mkl_dcsrmultd.</CODE></DIV><BR /><DIV>We really need the parallellism because we are operating on some 6000x6000 matrices and a single core is really slow.</DIV><BR /><DIV>Thanks a lot,</DIV></DIV><DIV>Ionut</DIV> Sat, 18 Jun 2011 06:10:45 GMT ionutg 2011-06-18T06:10:45Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Is-mkl-dbsrmm-parallel/m-p/803106#M3224 Hello everyone,<DIV><SPAN style="font-family: Verdana, Arial, Helvetica, sans-serif;"><BR /></SPAN> <DIV>I was wondering whether <CODE>mkl_dbsrmm</CODE> has a parallel implementation. I have set both <CODE>OMP_NUM_THREADS</CODE> and <CODE>MKL_NUM_THREADS</CODE> to 8 and I checked that number with<PRE>[fortran] nthr = mkl_domain_get_max_threads( MKL_BLAS )[/fortran]</PRE> </DIV><DIV></DIV><DIV>However, when I check the CPU usage with top it never goes above 100%. I remember seeing higher CPU usage (400%) when using <CODE>mkl_dcsrmultd.</CODE></DIV><BR /><DIV>We really need the parallellism because we are operating on some 6000x6000 matrices and a single core is really slow.</DIV><BR /><DIV>Thanks a lot,</DIV></DIV><DIV>Ionut</DIV> Sat, 18 Jun 2011 06:10:45 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Is-mkl-dbsrmm-parallel/m-p/803106#M3224 ionutg 2011-06-18T06:10:45Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Is-mkl-dbsrmm-parallel/m-p/803107#M3225 <P>Hello Ionut,<BR /><BR />I do seethe User Guide says it should be threaded, so we can check if that is accurate. The scaling will depend on the sparsity pattern so it's possibleyour case is not optimal.Is performance pegged at 100% or just always below it?<BR /><BR />Todd</P> Fri, 24 Jun 2011 22:23:48 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Is-mkl-dbsrmm-parallel/m-p/803107#M3225 Todd_R_Intel 2011-06-24T22:23:48Z