ALPACA

Overview

ALPACA is an MPI-parallelized C++ code framework to simulate compressible multiphase flow physics. It allows for advanced high-resolution sharp-interface modeling empowered with efficient multiresolution compression. The modular code structure offers a broad flexibility to select among many most-recent numerical methods covering WENO/T-ENO, Riemann solvers (complete/incomplete), strong-stability preserving Runge-Kutta time integration schemes, level-set methods and many more.

Getting Started

Installation

Recursively cloning in the case of a fresh installation

git clone --recursive https://github.com/tumaer/ALPACA.git

or in the case of an existing download

git fetch && git submodule update --init --recursive

After which we first need to install ALPACA's dependencies, ALPACA depends on

• MPI
• HDF5

On clusters, the two are likely going to be available as module to load. Outside of such computing environment, we need to make sure that we have them available on our system.

MPI Installation Instructions

To install and setup MPI, we have the choice of using OpenMPI, and MPICH. This instruction here is for OpenMPI, but applies equally as much for MPICH. Creating the build directory:

mkdir mpi-build && export MPI_BUILD_DIR=$(PWD)/mpi-build

To then begin the installation of MPI, we first have to download the source:

wget https://download.open-mpi.org/release/open-mpi/v4.1/openmpi-4.1.5.tar.gz
tar -xzf openmpi-4.1.5.tar.gz && cd openmpi-4.1.5

We then have to configure our installation, and compile the library:

./configure --prefix=$MPI_BUILD_DIR
make -j && make install

After which we are left to export the MPI directories:

export PATH=$MPI_BUILD_DIR/bin:$PATH
export LD_LIBRARY_PATH=$MPI_BUILD_DIR/lib:$LD_LIBRARY_PATH

If your cluster environment comes with its own MPI library, you should always prefer using the system MPI library over doing a source install.

HDF5 Installation Instructions

To install HDF5, we roughly follow the same outlines as the ones for the MPI installation. Creating the build directory:

mkdir hdf5-build && export HDF5_BUILD_DIR=$(pwd)/hdf5-build
mkdir hdf5-install && export HDF5_INSTALL_DIR=$(pwd)/hdf5-install

To then begin the installation of HDF5, we have to get the source, and then unpack it:

wget https://support.hdfgroup.org/ftp/HDF5/releases/hdf5-1.8/hdf5-1.8.23/src/hdf5-1.8.23.tar.gz
tar -xzf hdf5-1.8.23.tar.gz && cd hdf5-1.8.23

Set the compilers to be the MPI-compilers:

export CXX=mpic++
export CC=mpicc

After which we have to configure our installation, and then compile the library:

cmake -GNinja -B ../hdf5-build/ -S . \
    -DCMAKE_INSTALL_DIR=$(pwd)/../hdf5-install \
    -DCMAKE_BUILD_TYPE=Release \
    -DCMAKE_C_COMPILER=$(pwd)/../mpi-build/bin/mpicc \
    -DCMAKE_CXX_COMPILER=$(pwd)/../mpi-build/bin/mpic++ \
    -DHDF5_ENABLE_PARALLEL=On \
    -DHDF5_BUILD_CPP_LIB=On \
    -DALLOW_UNSUPPORTED=On

To then build and install from the build directory

cd $HDF5_BUILD_DIR
ninja && ninja install

Having MPI & HDF5, we can then install ALPACA with

cmake -GNinja -B ../alpaca-build/ -S . \
    -DCMAKE_BUILD_TYPE=Release \
    -DCMAKE_C_COMPILER=mpicc \
    -DCMAKE_CXX_COMPILER=mpicxx \
    -DHDF5_DIR=$HDF5_INSTALL_DIR/cmake

to build, we then invoke CMake again

cmake --build ../alpaca-build/

We highly recommend using ccache together with CMake. To do so, add the following flags to the configuration step of CMake:

-DCMAKE_C_COMPILER_LAUNCHER=ccache
-DCMAKE_CXX_COMPILER_LAUNCHER=ccache

Testing

To validate the installation, we recommend running unit-tests after the completed installation. To do so

ninja Paco -j 4

after which we can run single-, as well as two-core tests to verify the correctness of the installation.

mpiexec -n 1 ./Paco [1rank]
mpiexec -n 2 ./Paco [2rank]

For further instructions, first steps, and API documentation, please consult the ReadTheDocs.

Academic Usage

If you use ALPACA in an academic setting, please cite our papers.

Acknowledgments

ALPACA has received support from multiple funding bodies over the course of its inception:

• This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program: ERC Advanced Grant No. 667483, Prof. Dr. Nikolaus A. Adams, "NANOSHOCK - Manufacturing Shock Interactions for Innovative Nanoscale Processes"
• This project has received computing time on the GCS Supercomputer SuperMUC at Leibniz Supercomputing Centre (www.lrz.de) from the Gauss Centre for Supercomputing e.V. (www.gauss-centre.eu).
• This project has received funding from German Research Foundation (DFG).
• This project has received funding from the Bavarian State Ministry of Science and the Arts through the Competence Network for Scientific High Performance Computing in Bavaria (KONWIHR).