Diffuse functions problem

Just Got Here

2:12:53 AM PDT - Thu, Aug 29th 2013

Hi when I add diffuse functions to my basis set, either 6-31+G(d,p) or 6-31++G(d,p), I get an error.

However, without diffuse functions, calculation is running normal.
Do you have any idea what to do?

Quote:ERROR

WARNING : Found 56 linear dependencies

S eigenvalue threshold:     1.00000E-05

Smallest S eigenvalue :     5.55602E-08

Largest  S eigenvalue :     9.23222E-06

!! The overlap matrix has  56 vectors deemed linearly dependent with

   eigenvalues:

5.56D-08 9.92D-08 1.14D-07 1.40D-07 1.55D-07 1.69D-07 1.88D-07 2.22D-07

2.59D-07 2.71D-07 3.01D-07 3.68D-07 4.35D-07 4.64D-07 4.69D-07 5.36D-07

6.02D-07 6.13D-07 6.83D-07 7.86D-07 8.69D-07 9.70D-07 1.07D-06 1.11D-06

1.18D-06 1.21D-06 1.53D-06 1.55D-06 1.67D-06 1.69D-06 1.83D-06 1.97D-06

2.13D-06 2.37D-06 2.67D-06 2.78D-06 2.93D-06 3.02D-06 3.13D-06 3.31D-06

3.43D-06 3.70D-06 3.88D-06 4.05D-06 4.38D-06 4.93D-06 5.45D-06 5.62D-06

6.17D-06 6.23D-06 6.81D-06 7.01D-06 7.21D-06 8.10D-06 8.50D-06 9.23D-06

     Superposition of Atomic Density Guess

     -------------------------------------

Sum of atomic energies:       -3812.93982290

     Non-variational initial energy

     ------------------------------

Total energy =   -3875.685546

1-e energy   =  -38099.241022

2-e energy   =   17714.439976

HOMO         =      -0.298597

LUMO         =      -0.033085

  Time after variat. SCF:  12836.0

  Time prior to 1st pass:  12836.0

Grid_pts file          = ./molecule.gridpts.0

Record size in doubles =  12289        No. of grid_pts per rec  =   3070

Max. records in memory =    148        Max. recs in file   = 126656358

Grid integrated density:     606.880615832235

Requested integration accuracy:   0.10E-05

5: WARNING:armci_set_mem_offset: offset changed -335503183872 to -335507378176

diis_driver: error creating ga        0

------------------------------------------------------------------------

------------------------------------------------------------------------

Forum Vet

10:15:17 AM PDT - Thu, Aug 29th 2013
Running out of global memory
Dpetrovic, You are running out of global memory. You can either a) increase the amount of global memory in the memory line of the input file or b) decrease the number of matrices stored for DIIS. I would recommend option b). You can implement it by typing the following lines in the DFT input section dft convergence diis 5 end The example above reduces the number of stored matrices from 10 to 5. More details at http://nwchemgit.github.io/index.php/Release62:Density_Functional_Theory_for_Molecules#CONVE... Cheers, Edo

Just Got Here

1:14:50 AM PDT - Fri, Aug 30th 2013
Thanks Edo, I'll try it

Forum >> NWChem's corner >> General Topics