Hello guys,
I am running a frequency calculation with NWChem on NeSI pan cluster but every time I increased the number of cores up to 60 cores or memory, some tasks are killed. Please help me. Find below my input file and the batch file. Thanks
echo
start AZE_BH3
title "Azetidine-borane mp2 calculation"
charge 0
geometry units angstroms print xyz
symmetry cs
N 0.63686 0.38325 0.00000
C 0.09002 -0.53377 1.05721
C -0.82112 -1.18206 0.00000
C 0.09002 -0.53377 -1.05721
H -0.36898 0.01458 1.87260
H 0.85953 -1.20859 1.42649
H -1.81646 -0.75277 0.00000
H -0.87546 -2.26517 0.00000
H -0.36898 0.01458 -1.87260
H 0.85953 -1.20859 -1.42649
H 1.64797 0.46806 0.00000
B -0.01774 1.86256 0.00000
H 0.38082 2.38875 1.01324
H -1.21447 1.67645 0.00000
H 0.38082 2.38875 -1.01324
end
memory stack 1500 mb heap 600 mb global 1500 mb
basis spherical
* library cc-pVTZ
end
mp2
# Exclude core electrons from MP2 treatment
freeze atomic
end
driver
clear
tight
end
task mp2 freq
- !/bin/bash -e
- SBATCH --job-name NWChem_job
- SBATCH --account uc00341
- SBATCH --time 2:00:00
- SBATCH --constraint avx
- SBATCH --tasks 10
- SBATCH --cpus-per-task 4
- SBATCH --tasks-per-node 5
- SBATCH --mem-per-cpu 8G
- SBATCH --output AZE_BH3_MP2_AVTZ3.out # Include the job ID in the names
- SBATCH --error AZE_BH3_MP2_AVTZ3.err # of the output and error files
- Since --output is set, there is no need to redirect output to a file
- on the NWChem command line.
export ARMCI_DEFAULT_SHMMAX=30000
module load OpenMPI/2.0.2-GCC-6.3.0
srun /projects/uc00341/nwchem/nwchem-6.8/bin/LINUX64/nwchem AZE_BH3_MP2_AVTZ3.nw
|