Intel® Math Kernel Library

linking pardiso


I'm trying to compile and run pardiso_solv_sym_inv.f90 program from PARDISO.
My compilation is succesfull but during the execution I get a segmentation fault error.
Obviously my linking in not correct.
Can anyone advice me how to link? Thanks.


My makerules file:

PROGRAMS = pardiso_solv_sym_inv

pardiso_solv_sym_inv : pardiso_solv_sym_inv.o
    $(FC) $(PFLAGS) -o $@ pardiso_solv_sym_inv.o $(PLIBS)

all : $(PROGRAMS)

    rm -v *.o *.mod $(PROGRAMS)

Size of tmp array for fgmres solver


From the Intel MKL reference manual for version 11.3, in the FGMRES interface description on p.1602, it is said:

tmp double array of size ((2*ipar[14] + 1)*n + ipar[14]*(ipar[14] + 9)/2 + 1)) used to supply the double precision temporary space for the RCI FGMRES computations

then on top of p.1603, there is a note saying

You can define this array in the code as double tmp[(2*ipar[14] + 1)*n + ipar[14]*(ipar[14] + 9)/2 + 1)] if you run only non-preconditioned FGMRES iterations.

Using MKL Library in C#

Hello, I am working on a project where the ultimate aim is to use the MKL library to perform a large number of matrix operations. Would it be more optimal for me to somehow try to import the necessary functions from the MKL library into C# directly or create my own custom DLL in C++ (using functions from the MKL library) and import that DLL into C#? Is there a time or efficiency gain in using either method?


No example for Cluster Sparse Solver in the Intel® Parallel Studio XE 2016 on Windows


I've installed the Intel Parallel Studio XE 2016 on my 64 bytes Windows 7 computer.

The new Cluster Sparse Solver is present in the mkl directory (include files mkl_cluster_sparse_solver.h,.f90). But I didn't find the examples for it (no archive in the mkl installation directory).

How do I obtain theses examples please?

Thanks in advance for your help.



Configuring C++ library to use MKL

I'm trying to configure Opt++ to use MKL on Debian 8. Their document ( indicates to use the syntax

./configure --with-blas=/path/to/blas/blaslibraryname.a

however, when I checked the following two paths, there are several .a files



Automated parallelizing not working with large matrixes

I have a code in fortran that calculates eigenvectors and eigenfunctions of matrix and process the results. I use the ifort with the -mkl compilation, the code runs in parallel for an square matrix (N,N) when N is less than 100 (more or less) but when I try larger matrixes N>300 it only runs as one thread. 

Any help?




Hi All

I often use older (and sometime new) fortran code that uses LINPACK, so I often replace those calls to LAPACK equivalent routines so that I may use MKL.  I have found that this can improve performance in some cases. This also allows easy use of Intel Phi coprocessors for large problems where offloading is warranted.

I was wondering if there exist a set of wrapper routines that in general will allow LINPACK to be replaced by MKL?



Having trouble referencing/creating MKL.DLL in Visual C# Studio

I am having a terrible time setting up Intel's Math Kernel for Visual C#. I have downloaded and ran the install file with no luck of accessing the Math Kernel Functions inside of the mkl.dll file. When I implement one of mkl.dll functions a BadImageFormatException is thrown. Online documentations instructs you to go through the visual studio command prompt and call:

    "mklvars.bat" ia32
    nmake ia32

When I run this throught the shell mklvars.bat ia32 is accepted and nmake ia32 throws and it throws an error within  the command prompt.

std::complex in zgemm


I modified an example from the intel website to use std::complex<double>  type with the zgemm function from MKL.

Unfortunately I am getting wrong results for complex matrix multiplication. I am wondering what I am doing wrong.

Thank in advance





Intel® Math Kernel Library abonnieren