Gaussian Job Scripts

From UFRC
Revision as of 14:25, 30 June 2017 by Moskalenko (talk | contribs) (Moskalenko moved page Gaussian SLURM to Gaussian Job Scripts)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

back to the Gaussian page The following is a sample job submission script for Gaussian. It loads "gaussian" module, which detects the CPU type of the machine on which your job runs and chooses the appropriate version of Gaussian for best performance. If OMP_NUM_THREADS is set, Gaussian will use it for the number of processors unless %nproc or %nprocshared is set in your Gaussian input file.

The number of threads specified to Gaussian via either method above must be consistent with the number of processors in your resource request i.e (--cpus-per-task=N).

Automatic Processor Detection

#!/bin/bash
#SBATCH --job-name=gaussianjob
#SBATCH --output=gaussian.out
#SBATCH --error=gaussian.err
#SBATCH --mail-type=ALL
#SBATCH --mail-user=YOUR-EMAIL-ADDRESS
#SBATCH --time=01:00:00
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=4    # number of CPU core to use
#SBATCH --mem-per-cpu=4000
#SBATCH --account=YOUR-GRUOP-NAME
#SBATCH --qos=YOUR-GROUP-NAME

module load gaussian

which g09

input=YOUR_g09_input.com
output=YOUR_OUTPUT.log

g09 < $input > $output