Geometry dependent GIAO NMR calculation error


Clicked A Few Times
I am running DFT GIAO NMR shielding calculations.

When the input is this, the calculation finishes fine:

echo

start molecule

title "isomerA2002"
charge 0

geometry units angstroms print xyz autosym
  C       -5.39320       -0.89060       -0.36820
H -5.13760 -1.73170 -1.02440
C -4.75950 0.41410 -0.93920
H -5.16810 1.25980 -0.36880
C -4.17510 -1.81160 3.68060
C -4.98620 -0.70560 3.43390
C -5.38610 -0.41000 2.12900
C -4.97900 -1.21600 1.05340
C -4.16610 -2.32960 1.31850
C -3.76640 -2.62410 2.62430
H -3.86520 -2.04290 4.69640
H -5.31210 -0.07480 4.25680
H -6.02970 0.44880 1.94880
H -3.83770 -2.97650 0.50800
H -3.13700 -3.48920 2.81710
O -6.82240 -0.80400 -0.35730
H -7.06730 -0.33200 -1.18220
C -5.16290 0.61780 -2.42610
H -4.75790 -0.20040 -3.03330
C -3.23080 0.41420 -0.79160
H -2.79330 1.35300 -1.14300
H -2.92500 0.31100 0.25400
H -2.78090 -0.40630 -1.36060
C -4.70340 1.95310 -3.00700
H -5.11230 2.09590 -4.01350
H -5.06900 2.78810 -2.39940
H -3.61480 2.01810 -3.07840
O -6.59070 0.55930 -2.54250
H -6.81150 0.72960 -3.47820
end

basis
 * library 6-31G**
end

dft
 xc b3lyp
mult 1
end

task dft energy

property
 shielding
end
task dft property

<continued>

Clicked A Few Times
But when I supply this input file, the calculation ends with an error. Input:
echo

start molecule

title "1confnwinp001"
charge 0

geometry units angstroms print xyz autosym
  C        3.52630       -0.35110       -0.06330
C 4.21410 0.82950 -0.33660
C 3.57420 2.05780 -0.18350
C 2.24540 2.10550 0.24370
C 1.53960 0.92420 0.52620
C 2.19730 -0.30440 0.36370
C 0.09550 1.00500 0.96520
C -0.85760 1.25120 -0.19780
C -2.13420 0.63010 0.31320
C -1.59400 -0.60720 1.02430
O -0.34630 -0.20440 1.61910
H -0.00530 1.80060 1.71440
H -0.52980 0.74510 -1.11560
H -0.96550 2.31260 -0.44360
H -2.63340 1.30660 1.01750
O -3.03090 0.30590 -0.73430
C -2.54700 -1.14440 2.09220
H -1.37930 -1.40120 0.29700
C -2.00160 -2.40790 2.76360
C -2.97030 -2.94700 3.80500
H -1.82050 -3.18120 2.00800
H -1.04130 -2.19480 3.24670
H 4.02380 -1.31040 -0.18180
H 5.24850 0.79210 -0.66860
H 4.11010 2.97930 -0.39680
H 1.76270 3.07430 0.35600
H 1.67650 -1.23700 0.57230
H -2.70140 -0.37960 2.86370
H -3.52220 -1.36050 1.63990
H -3.14950 -2.20880 4.59320
H -3.93200 -3.20480 3.35000
H -2.56220 -3.84940 4.27090
H -3.25040 1.13400 -1.19720
end

basis
 * library 6-31G**
end

dft
 xc b3lyp
mult 1
end

task dft energy

property
 shielding
end
task dft property

The DFT single point energy works fine, but the NMR calculation fails. The relevant part of the output:
Wrote CPHF data to ./molecule.shieldcphf

Calc. par tensor-> nonrel
------------------------------------------------------------------------
get_dia: nga_create failed g_h11 all 0
------------------------------------------------------------------------
------------------------------------------------------------------------
For more information see the NWChem manual at http://nwchemgit.github.io/index.php/NWChem_Documentation


For further details see manual section:                                                                                                                                                                                                                                                                


MPI_ABORT was invoked on rank 0 in communicator MPI COMMUNICATOR 4 DUP FROM 0
with errorcode -1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.


0:0:get_dia: nga_create failed g_h11 all:: -1
(rank:0 hostname:moley pid:14832):ARMCI DASSERT fail. ../../ga-5-2/armci/src/common/armci.c:ARMCI_Error():208 cond:0

Does anyone has any ideas what is going wrong here? Any help would be much appreciated.

Forum Vet
You are running out of global memory.
Please add the following line at the beginning of your input file

memory stack 600 mb heap 200 mb global 1000 mb

Clicked A Few Times
Thank you very much, it worked! Spent 2 days trying to figure this out.

I guess a little less cryptic error messages might have helped.


Forum >> NWChem's corner >> General Topics