# Test runs with coupling of LAMMPS and LATTE
Step 0: What LATTE currently supports
Step 1: Build LAMMPS
Step 2: Install or download/build MDI code coupling package
Step 3: Download/build LATTE
Step 4: Perform test runs in any of 3 modes
---------------------------------
---------------------------------
Step 0: What LATTE currently supports
LATTE can be used with fix mdi/qm to perform QM calculations of an
entire system, but not with fix mdi/qmmm to perform QMMM simulations.
LATTE can calculate a QM energy and virial as well as QM forces on
each atom.
LATTE does not support MPI parallelism, so run it on a single MPI
task. But it does support multi-threading via OpenMP. By default it
will use all the threads available on your compute node. Depending on
the size of the problem and your hardware, it may be faster to run on
fewer threads. Thus for a production run it is a good idea to do some
short runs first with different thread counts to test performance.
For example, these commands will set the OpenMP thread count (to 4)
before performing a run. Unsetting the environment variable will
revert to the default (use all available threads).
bash:
% export OMP_NUM_THREADS=4
% unset OMP_NUM_THREADS
(t)csh:
% setenv OMP_NUM_THREADS 4
% unsetenv OMP_NUM_THREADS
All the example log files in this directory were run with OMP_NUM_THREADS = 1
---------------------------------
---------------------------------
Step 1: Build LAMMPS
The MDI and molecule packags are needed. Copy the final LAMMPS
executable into the examples/QUANTUM/LATTE directory.
Traditional make:
% cd ~/lammps/lib/mdi
% python Install.py -m mpi
% cd ~/lammps/src
% make yes-mdi yes-molecule
% make -j mpi
% cp lmp_mpi ~/lammps/examples/QUANTUM/LATTE
CMake:
% cd ~/lammps
% mkdir build; cd build
% cmake -DPKG_MDI=yes -DPKG_MOLECULE=yes ../cmake
% make -j
% cp lmp ~/lammps/examples/QUANTUM/LATTE/lmp_mpi
---------------------------------
---------------------------------
Step 2: Install or download/build MDI code coupling package
To simply use the MDI python module for the example scripts in this
dir, you only need to install the MDI python module in your
environment. Alternatively, you can download/build MDI itself, which
will include its documentation and examples (as well as libraries and
Python interface).
NOTE: It should be fine to leave off the version number for the python
module installs, to just grab the latest MDI version. 1.4.16 is the
version of MDI LAMMPS is currently using.
(option 1) Install MDI python module via pip:
% pip install 'pymdi>=1.4.14'
(option 2) Install MDI python module via conda:
% conda install -c conda-forge pymdi=1.4.16
(option 3) Download/build MDI code coupling package
(a) clone the MDI Git repo
% git clone https://github.com/MolSSI-MDI/MDI_Library.git mdi
(b) build MDI
% cd mdi
% mkdir build; cd build
% cmake ..
% make -j
(c) Add a line like this to your ~/.bashrc or ~/.cshrc file so that
Python can find MDI:
For bash:
% export PYTHONPATH="${HOME}/mdi/build/MDI_Library:${PYTHONPATH}"
% hash -r
For (t)csh:
% setenv PYTHONPATH "${HOME}/mdi/build/MDI_Library:${PYTHONPATH}"
% rehash
(d) Check import of 3 Python modules which the script that wraps
LATTE will need:
% python
>>> import numpy as np
>>> from mpi4py import MPI
>>> import mdi
---------------------------------
---------------------------------
Step 3: Download/build LATTE
NOTE: As of March 2023, this ECP branch of the LATTE repo is the
latest developement version of LATTE to use with LAMMPS. It will be
become the default LATTE branch at some point in the future. This doc
page will be updated when that happens.
(a) clone the ECP branch of the LATTE Git repo
% git clone -b ECP https://github.com/lanl/LATTE.git latte
(b) build LATTE as a shared library
% cd ~/latte
% mkdir build; cd build
% cmake -DBUILD_SHARED_LIBS=on ../cmake
% make -j # should produce liblatte.so in build
(c) Add a line like this to your ~/.bashrc or ~/.cshrc file so that
the liblatte.so file can be found:
For bash:
% export LD_LIBRARY_PATH="${HOME}/latte/build:${LD_LIBRARY_PATH}"
% hash -r
For (t)csh:
% setenv LD_LIBRARY_PATH "${HOME}/latte/build:${LD_LIBRARY_PATH}"
% rehash
---------------------------------
---------------------------------
Step 4: Perform test runs in any of 3 modes
These tests are in lammps/examples/QUANTUM/LATTE
in.ch4 = AIMD with CH4 molecule
in.graphene = AIMD with graphene
in.series = series of 3 UO2 conformations
in.sucrose = AIMD of sucrose molecule
in.uo2 = two molecules of UO2
in.water = eight water molecules
** run LAMMPS and LATTE with TCP/IP, 1 proc each
lmp_mpi -mdi "-name LMP -role DRIVER -method TCP -port 8021" -log log.water.tcp.1 -in in.water &
python latte_mdi.py -mdi "-name LATTE -role ENGINE -method TCP -port 8021 -hostname localhost" latte.in.water
lmp_mpi -mdi "-name LMP -role DRIVER -method TCP -port 8021" -log log.uo2.tcp.1 -in in.uo2 &
python latte_mdi.py -mdi "-name LATTE -role ENGINE -method TCP -port 8021 -hostname localhost" latte.in.uo2
lmp_mpi -mdi "-name LMP -role DRIVER -method TCP -port 8021" -log log.series.tcp.1 -in in.series &
python latte_mdi.py -mdi "-name LATTE -role ENGINE -method TCP -port 8021 -hostname localhost" latte.in.uo2
** run LAMMPS and LATTE with MPI, 1 proc each
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.water.mpi.1 -in in.water : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.water
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.water.min.mpi.1 -in in.water.min : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.generic
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.uo2.mpi.1 -in in.uo2 : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.uo2
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.ch4.mpi.1 -in in.ch4 : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.generic
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.sucrose.mpi.1 -in in.sucrose : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.generic
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.graphene.mpi.1 -in in.graphene : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.generic
mpirun -np 1 lmp_mpi -mdi "-name LMP -role DRIVER -method MPI" -log log.series.mpi.1 -in in.series : -np 1 python latte_mdi.py -mdi "-name LATTE -role ENGINE -method MPI" latte.in.uo2
** run LATTE as plugin MDI engine, 1 proc
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.water.plugin.1 -in in.water.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.water.min.plugin.1 -in in.water.min.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.uo2.plugin.1 -in in.uo2.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.ch4.plugin.1 -in in.ch4.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.sucrose.plugin.1 -in in.sucrose.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.graphene.plugin.1 -in in.graphene.plugin
lmp_mpi -mdi "-name LMP -role DRIVER -method LINK -plugin_path ${HOME}/lammps/examples/QUANTUM/LATTE" -log log.series.plugin.1 -in in.series.plugin
