[CP2K-user] [CP2K:13581] radix error with LS-SCF and large system

[hut... at chem.uzh.ch]

Hi

Add

EXTENDED_FFT_LENGTHS

to the &GLOBAL section.

regards

Juerg Hutter
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut... at chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp... at googlegroups.com wrote: -----
To: "cp2k" <cp... at googlegroups.com>
From: "Ada Sedova" 
Sent by: cp... at googlegroups.com
Date: 07/02/2020 02:08AM
Subject: [CP2K:13581] radix error with LS-SCF and large system

Hi All,

I have a large system (about 5K atoms, protein) that I am trying to run in LS-SCF. I have found that when I set MGRID too high (over about 300) I get this strange crash:

*******************************************************************************
 *   ___                                                                       *
 *  /   \                                                                      *
 * [ABORT]                                                                     *
 *  \___/                     Index to radix array not found.                  *
 *    |                                                                        *
 *  O/|                                                                        *
 * /| |                                                                        *
 * / \                                                      pw/fft_tools.F:292 *
 *******************************************************************************


 ===== Routine Calling Stack =====

            6 pw_grid_setup
            5 pw_env_rebuild
            4 qs_env_rebuild_pw_env
            3 qs_env_setup
            2 qs_init_subsys
            1 CP2K

I am on OLCF Summit, running 7.0 with DBCSR and GPU support. Also using threads and MPI. Here I am using about 24-54 nodes. 
When I set the grid a bit coarser, I have no problems converging an SCF in under 5 minutes.

Thanks for any help!


  
  -- 
 You received this message because you are subscribed to the Google Groups "cp2k" group.
 To unsubscribe from this group and stop receiving emails from it, send an email to cp... at googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/4e342f56-a00c-477b-8962-92f0f58b8ef1o%40googlegroups.com.