NWChem is a powerful high performance computational chemistry program which can handle a wide range of different types of computational chemistry problems.
Using NWChem on RCC Resources
NWChem requires the use of a parallel gnu module (
gnu) as well as the NWChem module
nwchem, in that order. Shown below are some examples for running both the serial and parallel versions of NWChem on the HPC.
We use an example from the tutorial. Log in to the submit node (
ssh [user]@hpc-login.rcc.fsu.edu) and make a directory
nwchem in your home directory.
Now create a file named
n2.nw, then open it in a text editor and add the following to it:
title "Nitrogen cc-pvdz SCF geometry optimization" geometry n 0 0 0 n 0 0 1.08 end basis n library cc-pvdz end task scf optimize
Note that there must be at least two spaces before each command in the
.nw file. Proper formatting is available in the tutorial. Finally execute the program by using
nwchem, loading the module files beforehand.
module purge module load gnu-mvapich2 module load nwchem cd ~/nwchem nwchem_mpich2 n2
Parallel on the Local Node
Using Intel Open MPI, the only differences between the serial and parallel versions is the use of the
srun -n 2 nwchem_mpich2 n2
This will allow for nwchem to run on two processors on the local node.
Parallel Using SLURM
Create a SLURM submit script named
nwchem_submit.sh in your nwchem directory, similar to the one described in the SLURM job submission tutorial:
#!/bin/bash #SBATCH -J "nwchem_test_job" #SBATCH -n 2 #SBATCH -p genacc_q #SBATCH --mail-type=ALL #SBATCH -t 00:01:00 module purge module load gnu-mvapich2 module load nwchem srun nwchem_mpich2 n2
Now submit the job using
$ sbatch nwchem_submit.sh