Holzer M (2021)
Publication Type: Conference contribution, Conference Contribution
Publication year: 2021
Event location: ICP Stuttgart University
URI: https://www.cecam.org/workshop-details/1070
In this lecture session, an
overview of waLBerla, pystencils and lbmpy will be provided. Herein,
pystencils functions as a code generation python package, suitable for
generating highly optimised CPU or GPU code from a symbolic description
of an update rule
applied to many data points. This is a typical setup for most stencils
codes. Since the symbolic representation heavily relies on the computer
algebra system SymPy, a small introduction to SymPy will be provided.
After understanding the workflow of pystencils,
an introduction to the lattice Boltzmann generation framework lbmpy will
be given. lbmpy can be seen as a toolbox that derives the update rule
for the lattice Boltzmann method in a generic symbolic way. Pystencils
then takes this symbolic representation to
generate highly optimised low-level code. Thus lbmpy builds an extension
to pystencils. The generated code can be used to run stand-alone
directly within the python frameworks or combined with existing C++
frameworks. As an example, the massively parallel multiphysics
solver waLBerla will be shown.
APA:
Holzer, M. (2021). Combining particle-based and continuum modelling in soft matter physics with ESPResSo, PyStencils, and LbmPy. In Proceedings of the ESPResSo Summer School. ICP Stuttgart University.
MLA:
Holzer, Markus. "Combining particle-based and continuum modelling in soft matter physics with ESPResSo, PyStencils, and LbmPy." Proceedings of the ESPResSo Summer School, ICP Stuttgart University 2021.
BibTeX: Download