B3LYP bulk relaxation exits at first SCF

Miguel Rivera miguel.riv... at gmail.com
Wed Nov 30 11:35:16 UTC 2016


Dear Conrad,

Many thanks for your reply. After some more investigation, I think you 
suspect correctly. I tried running the following example
https://www.cp2k.org/exercises:2015_ethz_mmm:mo_ethene
With the same result. I reduced to an H2 molecule and it finally worked. 
There must be something taking a lot of memory in the HF calculations 
(given that this example is not a hybrid).

As for the hardware, I am running cp2k.popt on 8 cores of 4 GB each with 
mpi parallelization. I also tried it on one core of 64 GB to no avail. 
Surely that should be plenty for a simple ethene molecule? I am concerned 
that at this rate the full might of my local computer cluster will not be 
enough for a periodic molecular calculation.

Could someone try my input example to see if there is an issue with my 
installation? On my machine it usually fails after a few seconds to a 
minute.

Kind regards,

Miguel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20161130/c03ea1ec/attachment.htm>


More information about the CP2K-user mailing list