hybrid functional revPBE0-D3 gets stuck with no errors outputted

[lar...@lbl.gov]

Hello,

I am trying to run a simulation using the hybrid functional revPBE0-D3. The simulation runs but doesn't output anything, including any error messages (I have attached here the files for the simulation). I have been able to simulate the same system using a meta-GGA functional with no issues, so I am pretty confident there is nothing intrinsically wrong with the system. I don't have a lot of experience using hybrid functionals, however, so I am not sure what may be causing this issue.

Thanks in advance!

Luis

[hut...@chem.uzh.ch]

Hi

I don't see an obvious mistake. Maybe it just takes that much longer?
I would start with a much smaller system and explore the CPU and
memory needs of your type of calculation.

regards

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

-----cp...@googlegroups.com wrote: -----
To: "cp2k" <cp...@googlegroups.com>
From: lar...@lbl.gov
Sent by: cp...@googlegroups.com
Date: 03/05/2019 09:18PM
Subject: [CP2K:11374] hybrid functional revPBE0-D3 gets stuck with no errors outputted

--
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 cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at https://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.


[attachment "revpbe0-d3.zip" removed by Jürg Hutter/at/UZH]

[Matt W]

I'd suggest starting from a converged GGA calculation. If you screen on that initial density you can get a big speed up (though rerun to check).

EPS_PGF_ORB is very tight, you could increase it a bit. EPS_SCHWARZ is also quite strict, you could increase it to 10^-7 or even -6 and then check convergence when you have a good wavefunction.

+ Juergs advice.

Matt

[Luis Ruiz Pestana]

Thanks for the pointers! I finally made it work starting from a previously converged meta-GGA calculation, using EPS_SCHWARZ 1.0E-6, and EPS_PGF_ORB 1.0E-15 (wouldn't work for higher values of this parameter, complaining that the Kohn-Sham matrix is not 100% occupied). Here are the timings:

Step     Update method      Time    Convergence         Total energy    Change

  ------------------------------------------------------------------------------

     1 OT DIIS     0.15E+00  555.7     0.00082967     -2039.1909073186 -2.04E+03

     2 OT DIIS     0.15E+00  354.5     0.00032946     -2039.4948432943 -3.04E-01

     3 OT DIIS     0.15E+00  355.1     0.00015380     -2039.5448765937 -5.00E-02

     4 OT DIIS     0.15E+00  354.9     0.00007709     -2039.5495586907 -4.68E-03

     5 OT DIIS     0.15E+00  355.2     0.00002709     -2039.5501113855 -5.53E-04

     6 OT DIIS     0.15E+00  355.2     0.00001723     -2039.5501685709 -5.72E-05

     7 OT DIIS     0.15E+00  356.3     0.00000431     -2039.5502010572 -3.25E-05

     8 OT DIIS     0.15E+00  354.9     0.00000239     -2039.5502031502 -2.09E-06

     9 OT DIIS     0.15E+00  354.8     0.00000112     -2039.5502038116 -6.61E-07

    10 OT DIIS     0.15E+00  355.0     0.00000048     -2039.5502039566 -1.45E-07

  *** SCF run converged in    10 steps ***

Btw, I also had to play a little bit with the MEMORY parameter because for values above ~2000 the simulation was crashing (e.g. WARNING: New heap segment map at 0x1004be00000 failed: Cannot allocate memory). I contacted my system administrator and mentioned that it may be an issue of compiling cp2k using craype-hugepages. I'm not sure this is an issue many people will have, but I wanted to document it.

Thanks!

Luis

--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/zT-OSyja9gY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.

To post to this group, send email to cp...@googlegroups.com.
Visit this group at https://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/d/optout.

--

Luis A. Ruiz Pestana, PhD

Postdoctoral Fellow

University of California, Berkeley

Lawrence Berkeley National Laboratory

Berkeley, CA 94720