[Pw_forum] about OMP_NUM_THREADS

Paolo Giannozzi giannozz at democritos.it
Fri Sep 25 17:53:41 CEST 2009


Lorenzo Paulatto wrote:

> Welcome Jason, and thank you for providing all the necessary  
> information from the beginning.

... with the exception of an appropriate subject :-)

> You are using  
> norm-conserving pseudopotential, I think 50-60Ry should be enough

for Ge and T, I think it is more than enough.

One may spare time by making first a quick-and-dirty structural
optimization, then refining it with higher cutoff and more k-points.
I would start with a much lower cutoff (25Ry for instance) and the
Gamma point only (maybe with a few more bands and small smearing:
if the system is close to a metallic state, it is easy to run into
trouble otherwise). For damped MD, a good choice of the time step
is important. I personally prefer BFGS to damped MD.

Paolo
-- 
Paolo Giannozzi, Democritos and University of Udine, Italy


More information about the Pw_forum mailing list