Memory demanding for Haydoc and Inversion of BSE

Deals with issues related to computation of optical spectra, in RPA (-o c) or by solving the Bethe-Salpeter equation (-o b). Includes local field effects, excitons, etc.

Moderators: Davide Sangalli, Daniele Varsano, andrea.ferretti, andrea marini, Conor Hogan, myrta gruning

Post Reply
sdwang
Posts: 224
Joined: Fri Apr 09, 2010 12:30 pm

Memory demanding for Haydoc and Inversion of BSE

Post by sdwang » Fri Aug 16, 2019 4:01 pm

Dear developers,
I am running BSE using Haydoc and Inversion of solver,and I have found the Inversion needs much more memories than Haydoc solvers. Is It true?
I have kept the all other parameters the same for H and I solvers.

Thanks!

shudong
S. D. Wang
IMU,HOHHOT,CHINA
E-mail: sdwang@imu.edu.cn

User avatar
Daniele Varsano
Posts: 2655
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Memory demanding for Haydoc and Inversion of BSE

Post by Daniele Varsano » Thu Aug 22, 2019 11:10 am

Dear Shudong,
yes this is the case. In Haydock there is a memory distribution among cpu's which is not present in the inversion technique.

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

Post Reply