[Wien] lapw0_mpi

[Chiung-Yuan Lin]

Regarding the question I post on Feb 27 about the errors of a large case on
lapw0_mpi, since no one has any comment on that yet, I guess I may need to
ask my questions in a slightly different way:

1. UG provdes a way to estimate the amount of memory required for a
particular number of atoms/unit cell. Does "# of atoms" means # of
inequivalent atoms or "physical" atoms?
2. Are there dimension parameters to be tuned for lapw0 for different
sizes of cases? In the case of lapw1, the parameters are NMATMAX and
NUME.
I wonder if I can tune such parameters in lapw0 to fix problems for large cases.

Thank you,
Chiung-Yuan