PyMPDATA-MPI

PyMPDATA + numba-mpi coupler sandbox (with a long-term goal of developing a pure-Python LES system)

In principle, PyMPDATA-MPI adapts to API of PyMPDATA to solve the following equation in an environment with multiple nodes: $$\partial_t (G \psi) + \nabla \cdot (Gu \psi)= 0$$

It is done with the help of MPI(Message Passing Interface). Every worker is responsible for computing its part of the decomposed domain.

1 worker

2 workers

3 workers

4 workers

Credits:

Development of PyMPDATA-MPI has been supported by the Poland's National Science Centre
(grant no. 2020/39/D/ST10/01220).

copyright: Jagiellonian University
licence: GPL v3

Design goals

• MPI support for PyMPDATA implemented externally (i.e., within PySuperDropletLES)
• hybrid threading (internal in PyMPDATA, in the inner dimension) + MPI (outer dimension) parallelisation
• portability across major OSes (currently Linux & macOS; no Windows support due challenges in getting HDF5/MPI-IO to work there)

Related resources

open-source Large-Eddy-Simulation and related software

Julia

• https://github.com/CliMA/ClimateMachine.jl/

C++

• https://github.com/microhh/microhh
• https://github.com/igfuw/UWLCM

C/CUDA

• https://github.com/NCAR/FastEddy-model

FORTRAN

• https://github.com/dalesteam/dales
• https://github.com/uclales/uclales
• https://github.com/UCLALES-SALSA/UCLALES-SALSA
• https://github.com/igfuw/bE_SDs
• https://github.com/pencil-code/pencil-code
• https://github.com/AtmosFOAM/AtmosFOAM
• https://github.com/scale-met/scale

Python/Cython/C

• https://github.com/CliMA/pycles