B3LYP bulk relaxation exits at first SCF

[Conrad]

Miguel,

Unfortunately HF is very heavy on memory and scales badly becuase you must 
store the results of the electron-repulsion integrals.
A nice option in CP2K is to use the Auxiliary Density Matrix Method (ADMM) 
which is an approximation that will reduce the number 
of integrals by removing the "less important" ones. Care must be taken to 
set it up, but it can make some large calculations feasible. 

Refer to:
Manuel Guidon, Jürg Hutter, and Joost VandeVondele
Journal of Chemical Theory and Computation *2010* *6* (8), 2348-2364
DOI: 10.1021/ct1002225

Best wishes,
Conrad
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20161130/6ef87415/attachment.htm>