ORCA

In particular implemented methods are: Semi-empirical: INDO/S, MNDO, AM1, PM3, NDDO/1; Hartee Fock theory (RHF, UHF, ROHF and CASSCF) all in direct, semidirect, or conventional mode, different RI approximations; DFT including a reasonably large number of exchange and correlation functionals including hybrid DFT and the most recent double hybrid functionals; High level single reference correlation models: CCSD(T), QCISD(T), CEPA, CPF (with and without RI, Local); High level ab-initio individual selecting multireference methods (MRCI, MRMP2, MRMP3, MRMP4, MRACPF, MRAQCC, SORCI, DDCI) for ground- and excited-states; Geometry optimization in redundant internal coordinates using analytical gradient techniques for all SCF methods as well as MP2; Excited state calculations via TD-DFT and CI-singles (CIS); Scalar relativistic ZORA, IORA and Douglas-Kroll-Hess (DKH) approaches, picture change effects, all-electron basis sets, effective core potentials; QM/MM interface to GROMACS and Double hybrid functionals including a fraction of nonlocal correlation.

When results obtained with ORCA are published the following citattion is required as minimum.

Neese, F. (2012) The ORCA program system, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2, 73–78.

However the users are required also to cite in addition the original papers that report the development and ORCA implementation of the methods used in users' research. The full list of papers is given in users manual, available online via following link:

https://orcaforum.cec.mpg.de/OrcaManual.pdf

The package is installed on PENZIAS. Users must load module for orca before use. Type

module load orca

Below is an example input file for water.

! B3LYP def2-SVP Opt
# My first ORCA calculation
*xyz 0 1
O        0.000000000      0.000000000      0.000000000
H        0.000000000      0.759337000      0.596043000
H        0.000000000     -0.759337000      0.596043000
*
%PAL nprocs 4 end

ORCA can run applications in thread parallel mode (OpenMP). Thus the line %PAL nprocs 4 end in above input file. The line instructs the ORCA wrapper to run different parts of ORCA modules on 4 cores. At HPC hyper treating is not supported. Consequently the number of cores requested in SLURM script (ntasks), must match the number of cores requested in input file. Below is the simple SLURM script which run the above calculation on 4 core on a single node.

#!/usr/bin/bash 
#SBATCH -J orca_OMP 
#SBATCH --nodes=1
#SBATCH --ntasks=4
#SBATCH -o output.%j
#SBATCH --mem 16GB

export OMP_NUM_THREADS=4

export OMP_STACKSIZE=1G
export KMP_AFFINITY=scatter

cd $SLURM_SUBMIT_DIR

/share/usr/apps/orca/5.0.3/bin/orca water.inp > output_water.out
 

In order to use cross node parallelism (MPI) users must use mpi parallel versions of the ORCA routines accessible from orca wrapper. It is very important to submit correct numbers on nodes, cores-per-node and ntasks. Otherwise orca wrapper will not start OpenMPI job. The list is given below. Consequently the ntasks and nodes must be adjusted accordingly. Note that penzias has 2 types of nodes 16 core/64GB and 32 core/192GB.

 orca_anoint_mpi
 orca_autoci_mpi
 orca_casscf_mpi
 orca_cipsi_mpi
 orca_cis_mpi
 orca_cpscf_mpi
 orca_eprnmr_mpi
 orca_grad_mpi
 orca_gtoint_mpi
 orca_mcrpa_mpi
 orca_mdci_mpi
 orca_mm_mpi
 orca_mp2_mpi
 orca_mrci_mpi
 orca_pc_mpi
 orca_plot_mpi
 orca_pnmr_mpi
 orca_pop_mpi
 orca_rel_mpi
 orca_rocis_mpi
 orca_scfgrad_mpi
 orca_scfhess_mpi
 orca_scf_mpi
 orca_soc_mpi
 orca_util_mpi
 orca_vpot_mpi

More information about orca use including some examples can be found on ORCA portal at: https://orcaforum.cec.mpg.de/portal.php