Questions tagged [scalapack]

ScaLAPACK is a library of high-performance linear algebra routines for parallel distributed memory machines. ScaLAPACK solves dense and banded linear systems, least squares problems, eigenvalue problems, and singular value problems.

The goals of the ScaLAPACK project are the same than the one’s of LAPACK, namely:

efficiency (to run as fast as possible),

scalability (as the problem size and number of processors grow),

reliability (including error bounds),

portability (across all important parallel machines),

flexibility (so users can construct new routines from well-designed parts),

and ease of use (by making the interface to LAPACK and ScaLAPACK look as similar as possible).

62 questions
15
votes
1 answer

cholesky decomposition ScaLapack error

I'm getting the following error and i'm not sure why. { 1, 1}: On entry to PDPOTRF parameter number 2 had an illegal value { 1, 0}: On entry to PDPOTRF parameter number 2 had an illegal value { 0, 1}: On entry to PDPOTRF…
pyCthon
  • 11,746
  • 20
  • 73
  • 135
8
votes
1 answer

Understanding Block and Block-Cyclic Matrix Distributions

In working with parallel decompositions of matrices, I'm familiar with a block distribution, where we have (say) 4 processes, each with their own subregion of the matrix: So for instance here we have the number of processes in a row (procrows)…
gsamaras
  • 71,951
  • 46
  • 188
  • 305
6
votes
1 answer

scalapack matrix diagonalization (pdsyevd)

I am writing a small test code for parallel matrix diagonalization using ScaLAPACK's divide-and-conquer algorithm PDSYEVD in C. However I am new to ScaLAPACK and looking at the source it appears a rather scary amount of variables to set for which I…
jaap
  • 5,661
  • 2
  • 20
  • 25
4
votes
1 answer

Can't compile package with intel MKL ScaLapack and Blacs

I am trying to compile python package https://github.com/jrs65/scalapy which wraps mkl scalapack and blacs libraries. Compilation is successful however when I try to execute test programs I get: Intel MKL FATAL ERROR: Cannot load symbol…
Baranas
  • 71
  • 1
  • 6
4
votes
1 answer

How to initialize two distinct blacs contexts?

I have a computer with nproc processors and I'd like to initialize two blacs grids, one of the dimension p x q = nprocs and one of the dimension 1 x 1. Assume MPI allready initialized and a routine finding good block sizes, the first grid is…
chris
  • 317
  • 2
  • 9
4
votes
1 answer

MKL ScaLapack problems

I am trying to run a simple Hello World (MKL) ScaLapack example from http://acts.nersc.gov/scalapack/hands-on/etc/pddttrdrv/pddttrdrv.c.html but I stumbled on a problem (I am using MPICH2 and my OS is Windows). When I run the code with the MPI flags…
user3371583
  • 191
  • 9
4
votes
1 answer

BLACS context value and multiple MPI communicators

I am trying to make some tests with BLACS/Scalapack (C interface, Intel MKL version) making use multiple MPI communicators: particularly, what I am trying to obtain is a set of BLACS contexts/grids corresponding (one-to-one) to a set of (disjoint)…
4
votes
2 answers

Solving a banded system from C using LAPACK's DGBSV

I have tried to search for a similar thread related to this topic, but nobody seems to care for banded systems (...). I am interested in solving a banded matrix using LAPACK/ScaLAPACK from a C code. First, I want to achieve a sequential solution…
Eduardo
  • 697
  • 8
  • 26
3
votes
1 answer

Cholesky with ScaLAPACK

I am trying to do a Cholesky decomposition via pdpotrf() of MKL-Intel's library, which uses ScaLAPACK. I am reading the whole matrix in the master node and then distribute it like in this example. Everything works fine when the dimension of the SPD…
gsamaras
  • 71,951
  • 46
  • 188
  • 305
2
votes
2 answers

converting gsl linear algebra for use in scalapack or other parallel matrix library

I have a code deeply embedded with GNU Scientific Library (GSL) matrix arithmetic, the main computation of this code is solving a large system of linear equations that takes a very long time in serial and with GSL and BLAS functions, is there a way…
pyCthon
  • 11,746
  • 20
  • 73
  • 135
2
votes
0 answers

segmentation fault when solving large system of linear equations using scalapack PDGESV

A parallel fortran code that solves a set of linear simultaneous equations Ax = b using the scalapack routine PDGESV fails (exiting with segmentation fault) when the no. of equations, N, becomes large. I have not identified the exact value of N at…
2
votes
2 answers

wrapping the ScaLAPACK functions in cython

I used lapack and blas library to implement matrix-inversion and multiply matrices in cython. Here is my code: import numpy as np cimport numpy as np cimport cython import ctypes cdef extern from "f2pyptr.h": void *f2py_pointer(object)…
Dalek
  • 4,168
  • 11
  • 48
  • 100
2
votes
2 answers

Subdividing Scalapack grid

I am trying to calculate the eigenspectrum for a number of large matrices using Scalapack, but rather than distribute each matrix across all 32 processes. I'd rather distribute each matrix across 4 processes and calculate 8 matrices in parallel. I…
Stein
  • 3,179
  • 5
  • 27
  • 51
2
votes
0 answers

LAPACK: Solving large periodic banded system of equations

I have to solve a large number of simultaneous equations (~1000s) to solve at every time step for a general mean curvature flow problem. The problem is defined over closed manifolds so the boundary condition is periodic. I am using…
hkrish
  • 1,259
  • 1
  • 10
  • 11
2
votes
2 answers

Fix arithmetic error in distributed version

I am inverting a matrix via a Cholesky factorization, in a distributed environment, as it was discussed here. My code works fine, but in order to test that my distributed project produces correct results, I had to compare it with the serial version.…
gsamaras
  • 71,951
  • 46
  • 188
  • 305
1
2 3 4 5