NPAR (Vasp) equivalent in CP2K

Satish Kumar satis... at
Sun Feb 8 01:19:04 UTC 2015

Thank you.

On Friday, February 6, 2015 at 5:33:14 AM UTC-5, Florian Schiffmann wrote:
> Hi,
> as there are no bands in CP2K there is no parallelization over bands. BTW, 
> it does not decrease the cost of the calculation but makes it scale better 
> in parallel. CP2K scales pretty good (given you got a good network) anyway 
> so you wouldn't gain anyway.
> Flo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the CP2K-user mailing list