Running replica exchange jobs on bridges2 GPU Jackson Penfield https://conferences.xsede.org/c/message_boards/find_message?p_l_id=&messageId=2775993 2021-07-23T20:32:01Z 2021-07-23T20:32:01Z

Hi,<br /><br />I am attempting to run my replica exchange jobs using NAMD on GPU nodes. Usually this involves mpirun or charmrun, but I can&#039;t seem to use either of these programs when loading namd/2.13-gpu.<br /><br />Can someone suggest how I might do this. My example job file is below.<br /><br />Thanks,<br />Jackson<br /><br /><br />#!/bin/bash<br />#SBATCH -N 1<br />#SBATCH --ntasks-per-node=40<br />#SBATCH -p GPU<br />#SBATCH -t 0:30<br />#SBATCH --gres=gpu:v100:8<br />#<br />set echo<br />set -x<br /># Load the needed modules<br />module load namd/2.13-gpu<br />cd $PROJECT/charmm-ccl5-r2/namd<br />echo $SLURM_NTASKS<br />#<br />mkdir output<br />for n in $( seq 0 $((${SLURM_NTASKS}-1)) ); do<br /> mkdir -p output/${n}<br />done<br />#<br /><br />charmrun $BINDIR/namd2 +p ${SLURM_NTASKS} +replicas ${SLURM_NTASKS} job0.conf +stdout output/%d/job0.%d.log

Jackson Penfield 2021-07-23T20:32:01Z