A framework to implement a multifrontal scheme on GPU architectures with OpenCL
Facolta di Ingegneria dei Sistemi, Corso di Laurea in Ingegneria Matematica, Politecnico di Milano
Politecnico di Milano, 2011
@article{bonelli2011framework,
title={A framework to implement a multifrontal scheme on GPU architectures with OpenCL},
author={BONELLI, F.},
year={2011},
publisher={Italy}
}
In this work we analyze an open-source multifrontal solver implementation (UMFPACK) and modify it to transfer the computation load on an OpenCL device, typically a GPU. To achieve this result the dbOpenCL library has been created, which allows a neat integration of OpenCL code into existent C or C++ code. An analysis and pro ling of both the original and the modi ed UMFPACK implementation is given while solving an example problem built with the LifeV nite element library.
September 27, 2011 by hgpu
