[CP2K-user] [CP2K:21862] Overestimation of cell volume (>20%) using WB97X_D3

[Frederick Stein]

Hi Alessandro,
You have the following options:
- Use less memory to prevent the out-of-memory event
- Apply the auxiliary density matrix method (ADMM) to reduce the 
computational costs 
(see https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/AUXILIARY_DENSITY_MATRIX_METHOD.html), 
set suitable auxiliary basis sets (BASIS AUX_FIT <ADMM_BASIS> in the KIND 
sections) from BASIS_ADMM_UZH .
- Apply the RI approximation 
(see https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/XC/HF/RI.html) 
and https://manual.cp2k.org/cp2k-2025_2-branch/methods/dft/hartree-fock/ri_gamma.html 
for further reference
Best,
Frederick
Alessandro Azzali schrieb am Freitag, 26. September 2025 um 10:48:00 UTC+2:

> Thank you again! I'm trying with this corrected input, but I get this 
> message and the calculation stops just before starting  "SCF WAVEFUNCTION 
> OPTIMIZATION". I'm using 112 cores 448 GB RAM with different settings: 56 
> MPI, 2 openMP, 32 MPI 4 openMP, 1 MPI 1 openMP (adjusting MEMORY each time) 
> but same results.
>
> HFX_MEM_INFO| Est. max. program size before HFX [MiB]:                     
> 380
>
> (This for  56 MPI, 2 openMP) 
> How can it be solved?
> Il giorno giovedì 25 settembre 2025 alle 22:18:31 UTC+2 abd el ali el 
> omrani ha scritto:
>
>> Try to add vdw correction it may help
>>
>> On Thu, Sep 25, 2025, 8:55 PM 'Alessandro Azzali' via cp2k <
>> cp... at googlegroups.com> wrote:
>>
>>> Thanks to both of you! I'm still trying to fix the &HF section but I'm 
>>> having problems with the truncated cutoff and/or memory.
>>> Il giorno giovedì 25 settembre 2025 alle 09:33:46 UTC+2 Jürg Hutter ha 
>>> scritto:
>>>
>>>> Hi 
>>>>
>>>> You want to use a hybrid functional but you don't have the mandatory 
>>>> definition of &HF in your input. 
>>>> Have a look for the correct defintions of hybrid functionals in the 
>>>> tests or example sections. 
>>>>
>>>> regards 
>>>> JH 
>>>>
>>>> ________________________________________ 
>>>> From: 'Alessandro Azzali' via cp2k <cp... at googlegroups.com> 
>>>> Sent: Wednesday, September 24, 2025 7:42 PM 
>>>> To: cp2k 
>>>> Subject: [CP2K:21851] Overestimation of cell volume (>20%) using 
>>>> WB97X_D3 
>>>>
>>>> Hello everyone, 
>>>> I am optimizing cell and geometry of a crystalline solid using WB97X-D3 
>>>> functional, starting from the XRD experimental measurement, to my surprise 
>>>> the calculation enormously overestimated the volume of the cell, while 
>>>> maintaining a certain geometry. I did various tests but I can't understand 
>>>> why and where I went wrong. Would anyone be kind enough to give me a hand 
>>>> or some advice? 
>>>>
>>>> I attach input, outputs, structure and experimental. 
>>>> Details: CP2K version 2024.1; method GAPW; functional WB97X_D3, basis 
>>>> set TZV2P-MOLOPT-HYB-GTH. 
>>>>
>>>> Thank you in advance for your help. 
>>>> Best regards, 
>>>>
>>>> Alessandro 
>>>>
>>>>
>>>> Fai crescere le giovani ricercatrici e i giovani ricercatori 
>>>>
>>>> con il 5 per mille alla Sapienza 
>>>>
>>>> Scrivi il codice fiscale dell'Università 80209930587 
>>>> Cinque per mille<https://www.uniroma1.it/it/node/23149> 
>>>>
>>>> -- 
>>>> 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... at googlegroups.com<mailto:
>>>> cp2k+uns... at googlegroups.com>. 
>>>> To view this discussion visit 
>>>> https://groups.google.com/d/msgid/cp2k/01c98375-9bd3-4b59-9f69-ffe947cbd8c8n%40googlegroups.com
>>>> <
>>>> https://groups.google.com/d/msgid/cp2k/01c98375-9bd3-4b59-9f69-ffe947cbd8c8n%40googlegroups.com?utm_medium=email&utm_source=footer>. 
>>>>
>>>>
>>>
>>> *Fai crescere le giovani ricercatrici e i giovani ricercatori*
>>>
>>> *con il 5 per mille alla Sapienza*
>>>
>>> Scrivi il codice fiscale dell'Università 
>>> *80209930587**Cinque per mille <https://www.uniroma1.it/it/node/23149>*
>>>
>>> -- 
>>> 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... at googlegroups.com.
>>> To view this discussion visit 
>>> https://groups.google.com/d/msgid/cp2k/dbbd4cc6-623c-4bd8-81b5-cbbd8f67c178n%40googlegroups.com 
>>> <https://groups.google.com/d/msgid/cp2k/dbbd4cc6-623c-4bd8-81b5-cbbd8f67c178n%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 visit https://groups.google.com/d/msgid/cp2k/7af4e1ec-6de1-4125-9dc6-4608b091d50dn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250926/48601ed1/attachment.htm>