Hi everyone,
I’m running fairly large (about 500+ basis functions) ROHF-CCSD(T)/cc-pVTZ energy calculations on a cluster with 64GB on 8 cores. The jobs take longer to complete than is allowed on our cluster, so they die usually in the AAB or ABB ijk iterations in CCTRIPLES. The scratch files are being written to a local disk, so I was wondering if it is possible to restart the calculations using the scratch files. I have already tried using file32 to run a restart calculation using the “restart_file” command however this was unsuccessful and the calculation started as normal.
Is there a way to restart these jobs from the CCTRIPLES section, and which scratch files and input command would I need?
(Just as a sidenote, there a more efficient way to run these calculations in regards to my memory/CPU usage or input file? The input file is given below.)
Thanks!
memory 56 Gb
molecule complex {
0 2
N -4.038612 -1.659144 0.553674
N -0.213610 2.671504 1.141009
N 2.521557 -1.928979 1.800696
C -3.503760 -0.734029 0.101024
C -2.837823 0.399617 -0.455846
C -1.712949 0.260430 -1.213154
C -0.077348 1.582883 0.754777
C 2.558081 -1.071373 1.025414
C 0.060856 0.278586 0.205722
C 1.215994 0.030924 -0.723610
C 2.576562 -0.005253 0.010167
H 2.707365 0.948226 0.546431
H -3.226864 1.382879 -0.188038
H -1.410867 -0.727131 -1.566105
H -1.302421 1.132439 -1.723931
H -0.282734 -0.543136 0.838442
C 3.737294 -0.210188 -0.967470
H 1.269795 0.828323 -1.480515
H 1.072029 -0.927329 -1.245168
H 4.696171 -0.230214 -0.434467
H 3.757023 0.612150 -1.695532
H 3.621705 -1.159355 -1.509205
}
set globals {
basis cc-pVTZ
scf_type df
freeze_core True
basis_guess cc-pVDZ
guess read
reference rohf
s_orthogonalization canonical
}
set cachelevel 4
set print 2
energy(‘ccsd(t)’)