Please find below some example job scripts. You can use these as templates for your jobs. If you want any more examples added, then please email Christopher Woods.

Amber pmemd.cuda single GPU job

#!/bin/bash -login
module add apps/amber-14

pmemd.cuda -O -i mdin -o mdout

Submit using, for example:

• sbatch -p gpu -A other_mvz -N 1 --ntasks-per-node=1 --gres=gpu:1 -t 15-0 run.slurm

Amber pmemd.cuda.MPI dual-GPU job

#!/bin/bash -login
module add apps/amber-14

mpirun -N 2 pmemd.cuda.MPI -O -i mdin -o mdout

Submit using, for example:

• sbatch -p gpu -A other_mvz -N 1 --ntasks-per-node=2 --gres=gpu:2 -t15-0 run.slurm

GROMACS multicore or multinode mdrun_mpi job

#!/bin/bash -login
module add apps/gromacs-4.6.7

mpiexec.hydra -psm -notunepme -dlb yes -bootstrap slurm mdrun_mpi

Submit for a single-node (16 core) job using, for example:

• sbatch -p cpu -A other_mvz -N 1 --ntasks-per-node=16 -t 15-0 run.slurm

Submit for a multi-node (e.g. 10 nodes, 160 core) job using, for example:

• sbatch -p cpu -A other_mvz -N 10 --ntasks-per-node=16 -t 15-0 run.slurm