Intel® Math Kernel Library

Starting with mkl

Hi mkl users! I've just installed mkl under fedora 9 and I have some codes that i would like to run using mkl libraries, so I have to know how to do it properly. I have tried the following example:

Program Teste

implicit none

integer n,i,incx,incy
parameter (n=4, incx=1,incy=1)
real x(n), y(n), res, sdot
external sdot

do i=1,n


print*, 'sdot =',res
open (unit=1, file='resultado', status='unknown')



Usage of vsrnggaussian() - very easy question

Dear FORTRAN experts

I am stuck at a very beginner problem with FORTRAN and I hope that you can help me. I want to get a zero-mean, unit-varience (sigma=1) Gaussian-distributed number. I found the function vsrnggaussian() which should do the trick, but I have no idea of how to implement it in my FORTAN code. I found an example ( but that seems to be not for FORTRAN.

differences between c standard library and MKL


I have some differences between the c standard library and mkl when using vdTanh and tanh.

double val = 0.020630512405774085;
double e1 = tanh(val);
double e2;

vdTanh(1, &val, &e2);

The results are different in the last 2 digits of the double. Any ideas why this happens?

I am trying to compare some code which was written without MKL to a code with MKL. Should I have such differences?



OpenSuse 10.3 and MKL

Hi all,

I have OpenSuse 10.3 on my laptop. I have recently installed latest version of MKL but when I try to compile MKL's examples, I get this error:

ld: BFD (GNU Binutils) (SUSE Linux) internal error, aborting at ../../bfd/reloc.c line 5285 in bfd_generic_get_relocated_section_contents

ld: Please report this bug.

Have you experienced this problem? Does MKL work with OpenSuSE or just with SuSE Enterprise version?



Pardiso: refinement contraction rate

Dear Pardiso users:

I got the following errors by using pardiso (with ilp64 interface):

The following ERROE was detected: -4
*** error PARDISO ( numerical_factorization) error_num= -112
*** error pardiso: iterative refinement contraction rate greater 0.9, interrupt
*** error pardiso: zero pivot

There are 38436 equations with 58210068 nonzeros. Parameters and call sentence are as follows:

Cholecky decomposition with PARDISO

Hi, everyone,

I have recently run accross a problem, which requires manipulations with large sparse matrices, in particular, it is desired to efficiently compute Cholesky decomposition of a symmetric matrix of the order >=30000. I checked that PARDISO can factorize such matrix relatively fast (at phase 22), but how do I exctract the factored matrix without actually solving the associated system of equations?



can not find libmkl_lapack.a on Mac

Dear all,

I am using intel MKL version 10.0.11 on my intel-Mac machine. I'm trying to link my program to static lapack library "libmkl_lapack.a' which by default should have been in directory "/Library/Frameworks/Intel_MKL.framework/Versions/10.0.11/lib/32". However, in the respective directory, there is no such a file at all, I have checked em64, but thesituationis the same. Could anybody let me know how I could fix this problem?



How frequencies are stored in x_out array of 1D FFT?

Could you please explain how frequencies are stored in x_out array of 1D FFT?
I pretty sure, that the real and imaginary parts of the zero frequency component are in x_out[0] and x_out[1];
Where are the real and imaginary parts of positive frequencies increasing in magnitude stored?
Where are the real and imaginary parts of negative frequencies increasing in magnitude stored?
And where is real and imaginary parts of the one aliased point that contains the most positive and the most negative frequency stored?

Thank you,

Performance info for different 2D FFT versions

I am performing 2D FFT on real data with dimensions 512x512. It is however for simplicitly implemented as a 1024x1024 complex FFT.

I have been searching the Inte pages for info on speed comparing all different 2D FFTs, real, complex, in-place, out of place.

Where can I find som doc on this? Or maybe someone can tell me this info in this thread?


订阅 Intel® Math Kernel Library