Hello everyone,
We are using the ground state optimized structure to calculate the non-adiabatic coupling values between the ground state and excited singlet states along with the NAC between the excited states. Previously, we have used the same input and script without any problem but now the same is giving an error “Not enough PEs in MemFok”.
Input:
$rem
EXCHANGE CAM-B3LYP
BASIS def2-SVP
CIS_N_ROOTS 4
CIS_TRIPLETS false
SET_ITER 50
CIS_DER_NUMSTATE 5
CALC_NAC true
GUI=2
$end
$derivative_coupling
0 is the reference state
0 1 2 3 4
$end
Output error:
Excited state 4: excitation energy (eV) = 3.4584
Total energy for state 4: -4354.32815241 au
Multiplicity: Singlet
Trans. Mom.: 0.0426 X -0.1246 Y 0.0154 Z
Strength : 0.0014900017
D(377) → V( 1) amplitude = -0.2829
D(380) → V( 2) amplitude = 0.7834
D(381) → V( 1) amplitude = -0.4282
D(381) → V( 3) amplitude = -0.2260
Calculating Relaxed Density
Q-Chem fatal error occurred in module 0, line 50, PE 0:
Not enough PEs in MemFok
mpirun noticed that process rank 0 with PID 17007 on node node34 exited on signal 6
(Aborted).
remove work dirs /scratch/qchem16992.0 – /scratch/qchem16992.0
rm -rf /scratch/qchem16992.0
rm -rf /scratch/qchem16992
bash script:
#!/bin/bash
#this is job name line
#PBS -N NAC
#this is execution directory
#PBS -d /home/
#this is out.log file
#PBS -o out.log
#this is error.log
#PBS -e err.log
#resources like nodes processors and wallclock
#PBS -l nodes=1:ppn=4
ulimit -a > t.log
cat $PBS_NODEFILE >temp4.nodes
NP=cat $PBS_NODEFILE|wc -l
cd $PBS_O_WORKDIR
module load compilers/openmpi/1.10
source /apps/codes/qchem/5.1/qcenv.sh
qchem -nt 4 ortho_S0_NAC.in > ortho_S0_NAC.out
If we use ‘-np’, the error is also the same. We really appreciate any help you can provide.