[CP2K-user] [CP2K:17162] Re: Relation between cell size and computational time/required memory

Mostafa Abedi abedimostafa755 at gmail.com
Tue Jun 14 12:42:16 UTC 2022


Dear Sam,
Many thanks for your answer. I've experienced the same  behavior. As you 
suggested, it's better to keep the cell size as small as possible. 

Thanks,
Mostafa 

On Tuesday, June 14, 2022 at 3:46:42 AM UTC-4 wave... at gmail.com wrote:

> Hi Mostafa
>
> Cell size directly affects your calculation times. If you double a side 
> for a cubic cell, I would naively expect scaling on the order of O(3) or as 
> the cube. It's a bit more complicated, but keep the cells as small as 
> possible. Note that there is no separate simulation box in cp2k: cell in 
> sold state physics sense equals cp2k simulation box. If you don't have a 
> repeating cell, add PERIODIC NONE.
>
> Kind Regards
>
> Sam
> On Monday, June 13, 2022 at 8:47:24 PM UTC+2 abedimo... at gmail.com wrote:
>
>> Hi Everyone,
>> I've been trying to get the potential energies and forces for a certain 
>> number of atoms at different mass densities, from high to very low, under 
>> PBC. I've realized that as the cell size gets larger (going to the lower 
>> densities) the computational time and particularly the required memory 
>> increases. Now I have a cell with the length of 60 Å and it can not be 
>> run because it asks a lot of memory. I wonder if there is any correlation 
>> between the cell size and computational time/required memory and if so, how 
>> it scales. I should say I could mange to run the case with the cell size of 
>> 50 x 50 x 50 Å. Thanks.
>>
>> Best,
>> Mostafa
>>
>

-- 
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/fc51069e-a06d-4388-85ee-cb6cf6f9b3a0n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20220614/15801ecd/attachment-0001.htm>


More information about the CP2K-user mailing list