Psi4 1.5/psi4 1.0

4x4.1.0.in.txt (10.3 KB)
4x4.1.0.out.txt (219.0 KB)
4x4.1.0.in.txt (10.3 KB)
4x4.1.5.out.txt (225.2 KB)
4x415inPSI4.o897707.txt (1.4 KB)

I am having problems going from psi4 1.0 to psi 1.5

Per the attached files, using the same input deck, this runs successfully under 1.0, but fails using 1.5. The same queue parameters were used in both submissions; 16 nodes, 1000mb per node.

The log for the 1.5 run is appended below (as well as attached). This looks like a memory issue,but using a system that is half the size of the attached gives the same error. What am I doing incorrectly?

setting dmc scratch directory

============================================================
===== Summary of your PSI4 job =====

The input file is: 4x4.1.5.in
The time limit is 150:00:00 HH:MM:SS
The target directory is: /home/aubtxe/psi4/cellulose_sapt
The working directory is /scratch-local/aubtxe.4x415inPSI4.897707
The memory limit is: 16gb
The job will start running after: 2022-11-26T07:20:58
Job Name: 4x415inPSI4
Virtual queue: medium
QOS: --qos=medium
Constraints: --constraint=dmc
Using 16 cores on master node dmc34
Node list: dmc34
Nodes: dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34 dmc34
Command typed:
/apps/scripts/runpsi 4x4.1.5.in 4x4.1.5.out
Queue submit command:
sbatch --qos=medium -J 4x415inPSI4 --begin=2022-11-26T07:20:58 --requeue --mail-user=eldertj@auburn.edu -o 4x415inPSI4.o897707 -t 150:00:00 -N 1-1 -n 16 --mem-per-cpu=1000mb --constraint=dmc

Directory /scratch-local/aubtxe.4x415inPSI4.897707 already exists
psi4 -i 4x4.1.5.in -o 4x4.1.5.out -n 16
/var/spool/slurmd/job897707/slurm_script: line 66: 14851 Killed psi4 -i 4x4.1.5.in -o 4x4.1.5.out -n 16
slurmstepd: error: Detected 1 oom-kill event(s) in StepId=897707.batch. Some of your processes may have been killed by the cgroup out-of-memory handler.