[CP2K-user] The cutoff unit in CP2K

[Marcella Iannuzzi]

Coarser grids are used to collocate contributions to the density from 
smoother functions. 
The total density is then obtained on the finest grids using interpolation 
operators.
This operation can lead to differences depending on the choice of the 
grids. 
Please find more details in 
VandeVondele, J.; Krack, M.; Mohamed, F.; Parrinello, M.; Chassaing, T.; 
Hutter, J. Quickstep: Fast and Accurate Density Functional Calculations 
Using a Mixed Gaussian and Plane Waves Approach. *Comput Phys Commun* *2005*
, *167* (2), 103–128. https://doi.org/10.1016/j.cpc.2004.12.014.

Regards
Marcella

On Saturday, January 16, 2021 at 10:06:42 AM UTC+1 msha... at gmail.com 
wrote:

> Dear Marcella,
>
> Thank you very much for your response. I have another question about this, 
> what is the effect of NGRIDS keyword on the calculations. I see increasing 
> the NGRIDS increase the speed of calculations but there are small 
> differences in the total energy calculations. How does NGRIDS affect the 
> calculations?
>
> Thank you so much again.
> On Saturday, January 16, 2021 at 12:29:29 PM UTC+3:30 Marcella Iannuzzi 
> wrote:
>
>>
>> In GPW the electron density, and not the wave functions, is expanded in 
>> PW.
>> Therefore the PW cutoff is four time as large as a the one you would use 
>> to expand the wave functions, as it is done in Quantum ESPRESSO.
>> Whether or not the grids and PW parts of one GPW calculation become 
>> expensive and dominant depends on the system and on the method you are 
>> using. Hence increasing the PW cutoff, within a reasonable range, is often 
>> not too problematic.
>> Regards
>> Marcella
>>
>> On Saturday, January 16, 2021 at 8:46:10 AM UTC+1 msha... at gmail.com 
>> wrote:
>>
>>> Hello everyone,
>>>
>>> I wanted to ask a question about the cutoff value unit. The unit in the 
>>> manual is in Ry and the default value is 280 Ry and for some systems, we 
>>> may need to increase this value. I wanted to know why would we need such 
>>> huge values? In Quantum ESPRESSO 60-80 Ry is fine for some calculations and 
>>> adopting higher values is computationally expensive. Can you please explain 
>>> a bit more about the unit of the cutoff in CP2K? and why in CP2K a larger 
>>> value is required than QE?
>>>
>>> Thank you very much in advance.
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210116/41de1247/attachment.htm>