I am trying to run EOM-EE-CCSDT calculation for two roots of the same symmetry for CuO anion. I am using 2 Tb disk as scratch space, however, the calculation fails halfway through while calculating EOMCCSDT energies as the whole disk space is utilized (issue might be due to pagefile). Is there a way to know how much the disk space required is reduced using Cholesky? Or is there a way to use another disk if the disk that is initially used is fully utilized?
Well, you need to refer to the paper on the EOM-CCSDT implementation to figure out the savings with RI/Cholesky approaches. The feature to add another disk is unavailable.
I tried using Cholesky approach, my input looks like this:
$rem
BASIS = gen
SCF_GUESS = READ
CCMAN2 =1
CC_SAVEAMPL = 1
MEM_STATIC = 1000
MEM_TOTAL = 64000
ECP = gen
EE_STATES = [0,0,2,0]
GUI = 0
CHOLESKY_TOL = 4
JOB_TYPE = SP
METHOD = EOM-CCSDT
$end
I get an error in output file as:
Solving for EOMEE-CCSDT B1 singlet states.
Warning! Not enough singles to form guess subspace. Doubles guess vectors will be generated.
In case of poor EOM convergence, reduce requested EE_STATES.
Warning! Not enough singles to form guess subspace. Doubles guess vectors will be generated.
In case of poor EOM convergence, reduce requested EE_STATES.