CudneLB is a MPI+CUDA or MPI+CPU high-performance CFD simulation code, based on Lattice Boltzmann Method.
It provides a clear interface for calculation of complex physics, and implementing new models.
You'll need:
- R
- packages for R: optparse, numbers, rtemplate, gvector, polyAlgebra
- nVidia CUDA (if you want to use GPU)
- python, numpy (if you want to use the integrated python interpreter)
- python, sympy and R package: rPython (if you want to develop a model using python in place or R)
- MPI (e.g. OpenMPI)
You can install lot of these with the tools/install.sh script (if you are not afraid of running a script with sudo):
sudo tools/install.sh cuda 6.5-14 # only if your GPU supports cuda
sudo tools/install.sh r
sudo tools/install.sh openmpi
tools/install.sh rdep
sudo tools/install.sh python-dev
tools/install.sh rpython
sudo tools/install.sh module # only on CentOS
The install.sh
script is designed to work on Ubuntu (e.g. on the Travis-CI VMs).
The install.sh
script should work on CentOS.
You can install the sudo
parts by yourself, and use script to install R packages: rdep and rpython.
This should work:
module load mpi/openmpi-x86_64 # only on CentOS
make configure
./configure --enable-double --enable-graphics --with-cuda-arch=sm_20
# only CPU ./configure --enable-double --disable-cuda
make d2q9
This should also work:
CLB/d2q9/main example/flow/2d/karman.xml
Author: Łukasz Łaniewski-Wołłk
Co-authors:
Contributors:
Developed at: C-CFD Group at Warsaw University of Technology from 2012
This software is distributed under the GPL v3 License.
If you need this software under a different license, please contact the main author.
Contact: llaniewski(monkey)meil.pw.edu.pl