There are 4 queues/partitions to submit a job. They are TEST, CPU, GPU and LONG. User needs to create a job script as follows for submitting a job. The command sbatch is to submit job to scheduler and returns a job id. This job id can be used later for monitoring and managing jobs.

CPU partition script

Submitting job on one node

Job Submit script : sbatch submit_job_1node.sh

Sample script link: CPU partition script for N1

CPU partition script

Submitting job on four node

Job Submit script : sbatch submit_job_4nodes.sh

Sample script link: CPU partition script for N4

GPU partition script

Job Submit script : sbatch submit.sh

Sample script link: GPU partition script

Long partition script

Job Submit script : sbatch submit.sh

Sample script link: Long partition script

Test partition script

Job Submit script : sbatch submit.sh

Sample script link: Test partition script

If users want to run in the guassian module, they can use the below code in their script line.

export g16root=/opt/apps/gaussian

export GAUSS_SCRDIR=/home/kumar/tmp

source $g16root/g16/bsd/g16.profile

g16 filename.com

Sample script link: Gaussian script