-
Notifications
You must be signed in to change notification settings - Fork 6
Compile DFTB with lammps support
This page documents the compilation of DFTB+ with lammps support in Xsede-supermic cluster http://www.hpc.lsu.edu/docs/guides.php?system=SuperMIC
Currently Loaded Modulefiles:
- INTEL/14.0.2
- intel/14.0.2
- mvapich2/2.0/INTEL-14.0.2
- download the most recent version of lammps (lammps-11Aug17) from http://lammps.sandia.gov/download.html
- [qiangz@smic1]$ tar -xf lammps-stable.tar.gz
- [qiangz@smic1]$ cd lammps-11Aug17/src
- [qiangz@smic1 src]$ make mpi
This will end up with a compiled executable called lmp_mpi in src directory
To make it work with dftb+, we also need to compile a static library in /src by ivoking
- [qiangz@smic1 src]$$ gmake mpi mode=lib
which will return two library files called liblammps.a and liblammps_mpi.a
- one needs to go to example/COUPLE/fortran_dftb
[qiangz@smic1 src]$ cd ../examples/COUPLE/fortran_dftb/
- modify the makefile as follows
FC = mpif90 # replace with your Fortran compiler
CXX = mpic++ # replace with your C++ compiler
$(FC) $(FFLAGS) simple.o -o simpleF.x liblammps_fortran.a $(LAMMPS_SRC)/liblammps_mpi.a -lstdc++ /home/qiangz/impi-5-fftw3/lib/libfftw3.a- [qiangz@smic1 fortran_dftb]$ make
This will generate the following new files
- liblammps_fortran.a
- liblammps_fortran.so
- lammps.mod
The modified DFTB+ package (_dftb+mvapich.tar) was kindly offered by Dr Nir Goldman from LLNL.
- tar -xf dftb+_mvapich.tar
- cd dftb+_mvapich
Take a look at the README file before compilation, there would be a few files which one needs to change.
In Makefile.objdir
cp /home/qiangz/lammps-11Aug17/examples/COUPLE/fortran_dftb/lammps.mod $(OBJDIR)In sysmakes/make.x86_64-linux-ifort
FXX = mpif90
FXXOPT = -O2 -ip # -openmp
# Library paths
MKL_LIBDIR = $(MKLROOT)/lib/intel64
LAMMPS_DIR = /home/qiangz/lammps-11Aug17/examples/COUPLE/fortran_dftb
LAMMPS_DIR2 = /home/qiangz/lammps-11Aug17/src
LIB_BLAS = -L$(MKL_LIBDIR) -Wl,--start-group \
-lmkl_scalapack_lp64 -lmkl_intel_lp64 \
-lmkl_intel_thread -lmkl_core \
-lmkl_blacs_intelmpi_lp64 -lmkl_blacs_intelmpi_ilp64 \
-Wl,--end-group
LIB_LAPACK =
LIB_LAPACK += $(LAMMPS_DIR)/liblammps_fortran.a $(LAMMPS_DIR2)/liblammps_mpi.a #liblammps_mvapich.a
# Any other libraries to be linked
OTHERLIBS = -liomp5 -lstdc++ -lpthread /home/qiangz/fftw-3.3.6/lib/libfftw3f.a
- $ cd prog_dftb/
- $ make
The successful compilation will generate an executable in _src/prg_dftb/obj_x86_64-linux-ifort/dftb+
Dr. Goldman provides an example of running a short MD configuration of water under ambient conditions using mio-1-1 parameter set (see _dftb+mvapich/example)
The parameter set could be downloaded from http://www.dftb.org/parameters/download/mio/mio-1-1-cc/
-
wget http://www.dftb.org/fileadmin/DFTB/public/slako/mio/mio-1-1.tar.xz
-
cd example
-
adjust the path of your parameter set in dftb_in.hsd
Prefix = "/home/qiangz/dftb+_mvapich/mio-1-1/"-
run the test by
$ mpirun -np 2 ~/dftb+_mvapich/src/prg_dftb/_obj_x86_64-linux-ifort/dftb+
When running dftb+ calculation, it appears that dftb+ could be successfully invoked. However, the calculation stopped at the end of the 1st cycle of electronic calculation when it is supposed to communicate with lammps.
================================================================================
==
== DFTB+ (Density Functional Tight Binding and much more)
==
== Unofficial release (r4423)
==
== (ParserVersion = 4)
==
================================================================================
********************************************************************************
** Parsing and initializing
********************************************************************************
Interpreting input file 'dftb_in.hsd'
--------------------------------------------------------------------------------
*** Converting input from version 3 to version 4 ...
*** Done.
Reading SK-files:
/home/qiangz/dftb+_mvapich/mio-1-1/O-O.skf
/home/qiangz/dftb+_mvapich/mio-1-1/O-H.skf
/home/qiangz/dftb+_mvapich/mio-1-1/H-O.skf
/home/qiangz/dftb+_mvapich/mio-1-1/H-H.skf
Done.
Processed input in HSD format written to 'dftb_pin.hsd'
Starting initialization...
--------------------------------------------------------------------------------
PGRID:ORBITAL: 1 x 1 x 2
PGRID:ALLPROC: 1 x 2
PGRID:ATOM: 1 x 2
RECLATPOINTS: 73
Mode: MD without scaling of velocities
(a.k.a. NVE ensemble)
Self consistent charges: Yes
SCC-tolerance: 0.100000E-05
Max. scc iterations: 4
Ewald alpha parameter: 0.146801E+00
Spin polarisation: No
Nr. of up electrons: 128.000000
Nr. of down electrons: 128.000000
Periodic boundaries: Yes
Diagonalizer: Standard
Maximal SCC-cycles: 4
Time step: 0.826827E+01
Temperature: 0.943711E-03
Random seed: 12345
Initial charges: Set automatically (system chrg: 0.000E+00)
Included shells: O: s, p
H: s
K-points and weights: 1: 0.000000 0.000000 0.000000 1.000000
Extra options:
LAMMPS (11 Aug 2017)
Reading data file ...
triclinic box = (0 0 0) to (9.855 9.855 9.855) with tilt (0 0 0)
1 by 1 by 2 MPI processor grid
reading atoms ...
96 atoms
run 10 upto
Neighbor list info ...
update every 1 steps, delay 10 steps, check yes
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 0
ghost atom cutoff = 0
binsize = 9.855, bins = 1 1 1
0 neighbor lists, perpetual/occasional/extra = 0 0 0
Setting up Verlet run ...
Unit style : real
Current step : 0
Time step : 0.2
RECLATPOINTS: 73
iSCC Total electronic Diff electronic SCC error
1 -0.13254394E+03 0.00000000E+00 0.81936451E+00
2 -0.13259454E+03 -0.50593002E-01 0.62477159E+00
3 -0.13266329E+03 -0.68756719E-01 0.34144446E-01
4 -0.13266348E+03 -0.18701402E-03 0.18185590E-01
scc total CPU: 0.08 Wall: 0.08
Total DFTB Energy: -130.6808254718 H
Total Mermin free energy: -130.6808254718 H
TS value: 0.0000000000
force total CPU: 0.10 Wall: 0.10
Volume: 0.645902E+04 au^3
Pressure: -0.496936E-04 au -0.146204E+10 Pa
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
dftb+ 0000000000C62F89 Unknown Unknown Unknown
dftb+ 0000000000C6185E Unknown Unknown Unknown
dftb+ 0000000000C0FC62 Unknown Unknown Unknown
dftb+ 0000000000BC19D3 Unknown Unknown Unknown
dftb+ 0000000000BC658B Unknown Unknown Unknown
libpthread.so.0 000000338740F7E0 Unknown Unknown Unknown
dftb+ 00000000007693E5 Unknown Unknown Unknown
dftb+ 00000000007897C1 Unknown Unknown Unknown
dftb+ 0000000000B42809 Unknown Unknown Unknown
dftb+ 0000000000B2DA9A Unknown Unknown Unknown
dftb+ 0000000000AAE169 Unknown Unknown Unknown
dftb+ 0000000000AA7955 Unknown Unknown Unknown
dftb+ 0000000000AA8701 Unknown Unknown Unknown
dftb+ 0000000000763CC2 Unknown Unknown Unknown
dftb+ 0000000000464A4C Unknown Unknown Unknown
dftb+ 0000000000433256 Unknown Unknown Unknown
dftb+ 00000000004331E6 Unknown Unknown Unknown
libc.so.6 0000003386C1ED1D Unknown Unknown Unknown
dftb+ 00000000004330D9 Unknown Unknown Unknown
===================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= PID 7314 RUNNING AT smic1
= EXIT CODE: 174
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================