[CP2K-user] [CP2K:21591] Re: Runtypes ENERGY_FORCE and MD

[Nauman Hafeez]

Thanks so much, Marcella. You were right, the problem was not with the 
convergence. It was my cluster settings - memory to be exact.

Best regards,
Nauman
On Thursday, 26 June 2025 at 03:02:25 UTC-4 Marcella Iannuzzi wrote:

> Hi Neuman, 
>
> The SCF of the MD run did not even start, according to the output you 
> posted. This is not a convergence problem.
> Most probably this is a memory issue, but it is hard to tell with these 
> information. 
> Regards
> Marcella
>
> On Wednesday, June 25, 2025 at 5:25:07 PM UTC+2 hafeez... at gmail.com wrote:
>
>> Dear CP2K developers and group, 
>>
>> I have a system of two ions (Na+ and Cl-) in water, the total number of 
>> atoms are ~2,500. I run this system using 48 cpu nodes and 256g of memory. 
>> I've constructed the basis sets and pseudopotentials for the ions myself. 
>> My goal is to run an MD simulation. When I run the system for energy and 
>> forces the SCF converges very well simply using the OT setup with (mostly) 
>> default setting. When I switch to MD, I get an segmentation fault error.
>>
>> The MD ENSEMBLE I use is NVT with TIMESTEP of 0.5 fs, and TEMPERATURE of 
>> 300. I have tried changing the ENERGY_GAP to 0.1, choosing MINIMIZER as 
>> CG and LINESEARCH as GOLD in OT settings, but system either doesn't 
>> converge or I get the core dump error (Segmentation fault - invalid 
>> memory reference). 
>>
>> I was wondering how does the system converges with ENERGY_FORCE and not 
>> with MD - is it that the system is not duly optimized? Or choice of my 
>> basis set/pseudopotentials? I'm attaching the input script files for the MD 
>> I tried and the output log.
>>
>> Any help or guidance will be much appreciated. Thank you in advance. 
>>
>> Best,
>> Nauman
>>
>

-- 
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/7d1a15e7-3fdb-4d25-a8d0-1108914a56adn%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20250626/f0b6cc25/attachment.htm>