ABINIT Discussion Forums

Hi, all.

I have been trying to use the two parameters, gwcomp and gwencomp, to decrease the number of empty states in the GW calculation. I have found that if I set symchi to 1, there will be SIGSEV (11) error after entering the subroutine completechi0_deltapart. The change of symchi from 1 to 0 will fix the problem. Is this phenomenon related to my use of Coulom cut off for a low-dimensional system?

Another problem is on the sum rule monitored in the calculation. When I used a default gwencomp=2.0, I got messages like:
Average fulfillment of the sum rule on Im[epsilon] for q-point 2 : 2671.46 [%]
I wonder what the meaning of the fulfillment of sum rule larger than 100% is, although in the tests/paral/tV1.out the sum rule is like 344.19 [%]. Should I only take care of the worst fulfillment of sum rule to make sure all of them are larger than 100% ?

Thank you very much.

Regarding the first problem, I have just tested it in the abinit-6.6.1, and the running is OK with symchi = 1.