[Pw_forum] 5.0.1 "cannot allocate spsi" in cegterg

[Paolo Giannozzi]

On Jan 19, 2013, at 13:44 , Антон Литвиненко wrote:

> But than I tried to re-run older calculation with dithioamide of  
> 1,3-benzenedicarboxylic acid (80 atoms in unit cell) which  
> successufully had run on 4.3.2 version, and got the following:

are you sure you are re-running exactly the same calculation in the  
same conditions?
Memory usage hasn't increased recently AFAIK.

>      Error in routine  cegterg (5014):
>       cannot allocate spsi

there is not enough memory to run the calculation

> When I reduce k-point grid to gamma point only, the calculation  
> runs successfully.

of course: with gamma only you need to store just half of the  
coefficients of plane
waves. You may further reduce the memory by setting  
"diago_david_ndim=2", or
'diagonalization="cg"', as explained in the documentation, setting  
"mixing_ndim" to
4 or so.

P.
---
Paolo Giannozzi, Dept of Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222