Recent posts
https://software.intel.com/en-us/recent/545801
enZYTRI sequential?
https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/562267
<p>Hi<br />
I am testing the following code with composer_xe_2015</p>
<p>lwork = 256*n<br />
call ZSYTRF( 'U', n, afull, n, ipiv, work, lwork, error )<br />
call ZSYTRI( 'U', n, afull, n, ipiv, work, lwork, error )</p>
<p>using</p>
<p>gfortran -O3 -fopenmp read_blas.f90 -Wl,--start-group ${MKLROOT}/lib/intel64/libmkl_gf_lp64.a ${MKLROOT}/lib/intel64/libmkl_core.a ${MKLROOT}/lib/intel64/libmkl_gnu_thread.a -Wl,--end-group -ldl -lpthread -lm</p>
<p>After setting OMP_NUM_THREADS=16 I can see in "top" that ZSYTRF runs in parallel but ZSYTRI not. The matrix is large enough (n=15120)</p>
<p>Any idea how to solve this problem?</p>
<p>Olaf</p>
<p> </p>
<p> </p>
<p> </p>
<p> </p>
Fri, 17 Jul 15 07:26:50 -0700Olaf Schenk562267MKL: Partial factorization and Schur Complement?
https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/496283
<p>Hi,</p>
<p>I have a large symmetric sparse matrix (10000*10000) and I want to reduce the matrix to condense form of last N elements. In order to do that I need to calculate the schur complement form of the matrix:</p>
<p>K=[A B; C D]; S = A - B * inv(D) * C </p>
<p>K is symmetric sparse (10000*10000) ; S is condense form which is usually dense (N*N) N: between 1 to 1000.</p>
<p>My first approach was: 1. Calculate X = inv(D) * C with Pardiso sparse solver. 2. Calculate A - B * X with dgemm function of Math Kernel Library.</p>
<p>But I found the performance is very depend on N variable (size of condensed matrix) which in part1 is number of RHS and inpart 2 is the number of columns for X matrix.</p>
<p>If N = 9: Part1 time:0.06 seconds + Part2 time: 0.03 seconds = Total time around 0.09 seconds.</p>
<p>But if N=400: Part1 time:2 seconds + Part2 time: 3 seconds = Total time around 5 seconds.which is too much for my program!!!</p>
<p>Note: for both cases size of K is constant 10000*10000.</p>
<p>So it seems that Pardiso is not doing very well with so many RHS and dgemm matrix is not doing great with X having large number of columns.</p>
<p>I am using latest version of MKL. My program is in FORTRAN on Windows. Is there any way to speed up the process? </p>
<p>I found out that there are some libraries like MUMPS that have functions for partial factorization and calculating condense form of the sparse matrix. Is this feature available in Pardiso or Math Kernel library as well? If yes, would you please tell me the function name.</p>
<p>Thank you in advance for your help. </p>
Fri, 13 Dec 13 13:29:14 -0800vahid s.496283Lower triangular matrix of a symmetric sparse matrix
https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/495340
<p>Hi,</p>
<p>Is there any way to get lower (or upper) triangular matrix of a symmetric sparse matrix from pardiso solver? If not, what is the best way to do that using MKL intel in FORTRAN? I know I can get lower triangular matrix of a symmetric matrix with pbtrf() function but I could not find an function for sparse matrices!</p>
<p>The size of symmetric sparse matrix is around 10000*10000. </p>
<p>Thank you,</p>
<p>Vahid</p>
Fri, 06 Dec 13 15:37:21 -0800vahid s.495340PARDISO improved features
https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/348636
<p>Hello,</p>
<p>I don't think the following features are available as of today within the Intel MKL PARDISO, but I was wondering if it is planned to include them in future releases:</p>
<p>1- Schur complement for symmetric matrices, available in competitors such as MUMPS (page 9, <a href="http://graal.ens-lyon.fr/MUMPS/doc/userguide_4.10.0.pdf">http://graal.ens-lyon.fr/MUMPS/doc/userguide_4.10.0.pdf</a>) and PaStiX (page 34, <a href="https://gforge.inria.fr/docman/view.php/186/5707/pastixman.pdf">https://gforge.inria.fr/docman/view.php/186/5707/pastixman.pdf</a>)</p>
<p>2- distributed solver (available in University of Lugano PARDISO)</p>
<p>Thanks for your help.</p>
Mon, 17 Dec 12 04:11:58 -0800qweasd q.348636DSS vs. Pardiso
https://software.intel.com/en-us/forums/intel-math-kernel-library/topic/277966
<p>In different parts of my application, I need determinant, separate solves (331,333) and inverse of selected elements of a real sparse SPD matrix. If I understand correctly, only DSS computes determinant. Does it allow sparse RHS and separate solves (331-333)? What are my choices?Thanks in advance.MP</p>
Fri, 01 Jun 12 20:35:54 -0700sigmahat277966