Parallel Sparse Solver for Large Systems

Significant acceleration of the calculation using the Parallel Sparse Solver
Significant acceleration of the calculation using the Parallel Sparse Solver
The Parallel Sparse Solver is an equation solver optimized for sparse systems of equations. It can be used for all applications relating to statics and dynamics as well as the analysis of stability problems.

 

Using the Parallel Sparse Solver provides the following advantages over the standard equation solver:

  • Minimization of the amount of required memory space needed and the number of required computing operations
  • Significant speed advantage when solving the equation system
  • Exploitation of multiprocessor technology through parallel computing

To help demonstrate these enormous advantages, the required memory space and the calculation time using the Parallel Sparse Solver for a calculation example are shown below.

System properties
Sparse Solver
Nodes:276,320
Elements:350,925
Supports:844
Unknowns:1,657,920
Stiffness matrix:7.3 GB
Triangulation time:00:00:18 (h:m:s)

Library: Intel® Math Kernel Library, Copyright © Intel Corporation
Computer system: CPU Intel® i7-3770, 3.4 GHz, 32 GByte RAM

It is evident, that even enormous large equation systems can be solved in a short time with the Parallel Sparse Solver.

The advantages are even more obvious, if the system increases in terms of complexity or dynamic calculations are performed.