[CP2K-user] [CP2K:18189] POISSON WAVELET: Density non-zero on the edges & CPASSERT failed

[Augustin Bussy]

Hi Paolo,

it is generally not recommended to perform MP2 calculations with GPW integrals and all-electron basis sets. Because the integrals are evaluated numerically on a grid, dealing with the very localized core electrons would require a very fine grid. This translates to a large value of WFC_GPW%CUTOFF and a massive memory footprint. This is especially true if your simulation cell is large, as it is the case with you. For maximum efficiency and lowest memory usage, you should take a cell that is just large enough so that the WAVELET solver does not complain.

There are other types of integrals, but only GPW works for the calculation of gradients (I think). Therefore, a geometry optimization will require you to use this scheme. If you are interested in the geometry, I suggest you use basis sets from the BASIS_RI_cc-TZ file with pseudopotentials and the GPW scheme. I join a modified version of your original input file which runs without issues. Note that I relaxed the values of some of your EPS thresholds for better performance.

I hope that helps. Best,
Augustin
________________________________
From: cp2k at googlegroups.com <cp2k at googlegroups.com> on behalf of Paolo Sebastiano Floris <pa.floriss at gmail.com>
Sent: Tuesday, December 13, 2022 10:13 AM
To: cp2k <cp2k at googlegroups.com>
Subject: [CP2K:18184] POISSON WAVELET: Density non-zero on the edges & CPASSERT failed

Hello everyone,

I'm trying to perform a geometry optimization of an isolated molecule composed of 27 atoms and I get the warning

*** WARNING in pw/ps_wavelet_methods.F:241 :: Density non-zero on the ***
*** edges of the unit cell: wrong results in WAVELET solver ***

followed by " CPASSERT failed, mp2_ri_2c.F:1345 ".
I have looked at other threads such as
https://groups.google.com/g/cp2k/c/J3WxEhnuNOY/m/lJ1GMSYyAQAJ
https://groups.google.com/g/cp2k/c/g4eXkddU2TM/m/U72NxzdpBQAJ
but they don't solve my problem.

I attach my input and (part of the) output files: I'm using double hybrid DSD-PBEP86 (as detailed in https://pubs.rsc.org/en/content/articlelanding/2011/cp/c1cp22592h) which involves performing some MP2 calculations. You can see in the input that I have reduced the accuracy of this section (I get "out of memory" errors otherwise): could this be the cause of the error?

Thanks for the help!
Sincerely,
Paolo

--
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+unsubscribe at googlegroups.com<mailto:cp2k+unsubscribe at googlegroups.com>.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/a67ebc3e-63a3-42e7-8c40-d84e466eb9c6n%40googlegroups.com<https://groups.google.com/d/msgid/cp2k/a67ebc3e-63a3-42e7-8c40-d84e466eb9c6n%40googlegroups.com?utm_medium=email&utm_source=footer>.

-- 
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+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/GV0P278MB0370EE677F309E77040892E7D3E39%40GV0P278MB0370.CHEP278.PROD.OUTLOOK.COM.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20221213/1fe3bd11/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: monomer.inp
Type: application/octet-stream
Size: 3664 bytes
Desc: monomer.inp
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20221213/1fe3bd11/attachment-0001.obj>