[Pw_forum] Force calculation time in relaxation under electric field

[Paolo Giannozzi]

The calculation of forces in DFT+U may take a significant amount of time
when there are many atoms with U correction. Recent versions of QE are much
faster than earlier ones in this respect. Which version of QE are you
using, how many atoms (with and without Hubbard U) and on how many
processors are you running?

Paolo

On Mon, Sep 21, 2015 at 4:21 AM, Mostafa Youssef <myoussef at mit.edu> wrote:

> Dear all,
>
>
> I performed ionic relaxation for 3x3x3 unit cells of a semiconducting
> oxide under electric field (lelfield =.true.) using ultrasoft PP and 1x1x3
> MP K-point grid where 3 is in the direction of the electric field. These
> are GGA+U calculations. I noticed in the time report that force
> calculations time is becoming even more than the electrons time as shown
> below:
>
>      init_run     :      8.32s CPU      8.91s WALL (       1 calls)
>      electrons    :  42723.25s CPU  44995.21s WALL (       5 calls)
>      update_pot   :      2.62s CPU      3.19s WALL (       4 calls)
>      forces       :  66735.95s CPU  66943.64s WALL (       5 calls)
>
> This is first time in which I find PWscf use more time in force
> calculations than in electrons. For the same system and without electric
> field , forces time would be about 0.2 of electrons time.  Is this a normal
> behavior and is there a recommended possibility to reduce force calculation
> time?
>
> I attempted using Norm Conserving PP's but unfortunately I could not get
> the code pass the 1 ionic step because of the S matrix not being positive
> definite.
>
> I appreciate your thoughts on this.
>
> Regards,
> Mostafa Youssef
> MIT
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150921/13a2d036/attachment.html>