Create a submission script named sub_orca_5.0.1
as follows:
#!/bin/csh # Set name of input and number of CPUs per node # If no ncpus provided, default will be 1 set input=$1 if (null$2 == null) then set ncpus=48 else set ncpus=$2 endif # Set memory if (null$3 == null) then set ngb=32 else set ngb=$3 endif # Set time limit if (null$4 == null) then set nhour=168 else set nhour=$4 endif # Set number of nodes for the calculation if (null$5 == null) then set nnode=1 else set nnode=$5 endif # Get rid of the possible .inp extension if($input:r.inp == $input) set input=$input:r set ngb=`echo "scale=0;$ngb*1024 / $ncpus"|bc -l` echo "running calculations for $input" # Make the slurm file set wd=`pwd` echo "#\!/bin/sh" > $input.slurm echo "#SBATCH --nodes=$nnode --ntasks-per-node=$ncpus" >> $input.slurm echo "#SBATCH --mem=""$ngb""G" >> $input.slurm echo "#SBATCH --time=""$nhour"":00:00" >> $input.slurm echo "#SBATCH --account=def-account-name" >> $input.slurm echo "" >> $input.slurm echo " cd $wd" >> $input.slurm echo "" >> $input.slurm echo "module add StdEnv/2020 gcc/10.3.0 openmpi/4.1.1" >> $input.slurm echo "module add orca/5.0.1" >> $input.slurm echo "" >> $input.slurm echo " /cvmfs/restricted.computecanada.ca/easybuild/software/2020/avx2/MPI/gcc10/openmpi4/orca/5.0.1/orca $input.inp > $input.out" >> $input.slurm sbatch $input.slurm
Add the following line to your .bashrc
file:
export PATH="/path-to-sub_orca_5.0.1/:$PATH"
Check if you have the Intel Fortran compiler and load it:
module spider intel module spider intel/2023.2.1 module add StdEnv/2023 module add intel/2023.2.1
Then compile the Fortran code:
ifort -o trnrot.x trnrot.f
Run the Fortran code as follows:
./trnrot.x
This will prompt you for details like the total number of atoms, the software (ORCA or Gamess US), and the atom indices of the ring in clockwise order.
Create a NICS input file for ORCA with the following template:
! RHF NMR wB97X cc-pVTZ AutoAux RIJCOSX UseSym TightSCF %output Print[ P_Basis ] 2 Print[ P_MOs ] 1 end %maxcore 1024 %pal nprocs 40 end * xyz 2 1 [Your geometry goes here] H: 0.0 0.0 1.0 newgto S 1 1 100000 1 end newauxJKgto S 1 1 200000 1 end H: 0.0 0.0 -1.0 newgto S 1 1 100000 1 end newauxJKgto S 1 1 200000 1 end *
Submit the ORCA job with the following command:
sub_orca_5.0.1 file_name.inp 32 0 2 1
[32 processors, 0 maximum memory, 2 hours, 1 node]