[Pw_forum] Plane waves parallelization
ouuing at gmail.com
Mon Apr 5 03:53:14 CEST 2010
Generally, it is not necessary to use many k point for so many atoms.
I think it is enough to use K_points 1x1x1.
On Sun, Apr 4, 2010 at 9:18 PM, Dimpy Sharma <dimpy.sharma at tyndall.ie> wrote:
> Hi Quantum espresso users,
> I was trying to run a scf calculation for my system (518 atoms), I used 100
> nodes to simulate the system however it crashed after 3 iterations without
> any error message (consumed 883MB).I used K_Points 2x1x1.
> From the user guide , I was reading the parallelism in quantum espresso,I
> could not follow "pools" and plane wave parallelization, and how to do the
> K_Point sampling in mpi parallelization for my system.
> Thanks and regards!
> Pw_forum mailing list
> Pw_forum at pwscf.org
State key Laboratory of Superhard Materials, Jilin University, China
Email: liuhanyu08 at mails.jlu.edu.cn ouuing at gmail.com
More information about the users