Thesis or internship »Accelerate the linear solvers within MESHFREE «
Fraunhofer-Gesellschaft · Kaiserslautern, Rheinland-Pfalz, DE
The »Grid-free Methods« department at the Fraunhofer Institute for Industrial Mathematics ITWM focuses on the modeling and efficient simulation of complex in...
Job description
The »Grid-free Methods« department at the Fraunhofer Institute for Industrial Mathematics ITWM focuses on the modeling and efficient simulation of complex industrial problems for the design and optimization of technical processes and products.: The efficient solution of large linear systems of equations is a key challenge in numerous scientific and industrial applications – from numerical simulations to data-driven models. The aim of this master’s thesis is to systematically analyse Krylov subspace recycling methods for dense linear matrices and to optimize them with respect to their performance. The insights gained will then be implemented in a prototype and integrated into the simulation software MESHFREE. Be part of change - Literature review of existing Krylov subspace recycling techniques for changing linear systems - Implementation of the most promising approaches to solve the linear systems arising from MESHFREE point clouds, for example for the hydrostatic pressure equation outside of MESHFREE - Analysis and investigation of the results using performance benchmarking - Extension and integration of the best methods into MESHFREE - Presentation of the results in research semi...