VSCF crash with "util jacobi: ierr 1"


Click here for full thread
Clicked A Few Times
Hi all,

Anyone know more about this error statement? I looked up the code behind util_jacobi and don't understand how ierr=1 is even possible.

For context, I'm trying to add anharmonic contributions into my IR calculation for a molecule with 41 atoms. At run time, my log file eventually stated "Starting diagonal potential on a grid of 16 points for 117 normal modes" It made it through a little over 100 normal modes find. Starting with normal mode twelve, it failed to converge. Normal mode 11 was the same story. The first grid point of normal mode 10 (the last positive frequency normal mode) and then threw an error message beginning with the text below. I thought the error may be related to the failed convergence, so I added the keyword cgmin to the DFT block, increased the max iterations and restarted. Unfortunately, the job immediately throws the same error. I'll attach my input script below the error message.

  • -*-*-*-*
warning !
*-*-*-*-*
SCF HAS NOT CONVERGED at this vscf point!


VSCF: Energy and dipole for grid point  1 along mode 10

                                NWChem DFT Module
-----------------




Summary of "ao basis" -> "ao basis" (spherical)
------------------------------------------------------------------------------
Tag Description Shells Functions and Types
---------------- ------------------------------ ------ ---------------------
H 3-21g 2 2 2s
C 3-21g 5 9 3s2p
Si 3-21g 7 13 4s3p


 Caching 1-el integrals
------------------------------------------------------------------------
util_jacobi: ierr 1
------------------------------------------------------------------------
------------------------------------------------------------------------
current input line :
7: task dft vscf
------------------------------------------------------------------------
------------------------------------------------------------------------
This error has not yet been assigned to a category
------------------------------------------------------------------------
For more information see the NWChem manual at
http://nwchemgit.github.io/index.php/NWChem_Documentation


For further details see manual section:
No section for this category



0:0:util_jacobi: ierr:: 1
(rank:0 hostname:n1293.becn pid:19172):ARMCI DASSERT fail. ../../ga-5.6.5/armci/src/common/armci.c:ARMCI_Error():208 cond:0
util_jacobi: ierr                    1
------------------------------------------------------------------------

Here's my input. Note that I don't think the driver block is doing anything:
echo

start excitingIR

scratch_dir /scratch

memory stack 60 heap 2 global 60 gb noverify

geometry units angstrom noautoz nocenter noautosym
 load GroundBreakingMolecule.xyz
end

BASIS "ao basis" spherical PRINT
 H library 3-21g
C library 3-21g
Si library 3-21g
END

dft
 xc b3lyp
odft
mult 1
mulliken
direct
maxiter 200
end

driver
 xyz xyzfiles/GroundBreakingMolecule
maxiter 400
end

freq
 temp 3 298.15 373.15 473.15
animate
end

task dft vscf