Hybrid parallelism for problems of hemodynamics: Freefem++ platform
talk by Kirill Pichon GOSTAF
We present a hybrid parallelization framework to solve large scale problems of hemodynamics in Freefem++ environment. Nowadays, multi-core processors became a universal standard and are widely exploited at all levels of high-end workstations and scientific clusters. We demonstrate that computational performance of a finite element code could be dramatically improved, when a proper use of a multi-core architecture is carried out. On 8-core shared memory machine, a hybrid parallelization approach performs considerably faster than a thread based BLAS implementation. We present a roadmap to solve a 6.5 million degrees of freedom problem using an open source Freefem++ platform, particularly its MPI version.
Three-dimensional model of an arterial bifurcation with a saccular aneurysm is studied. All numerical data is available online for public use.
|
