Re: parallelization for find-k calculations

On Jun 18, 2009, at 7:48 AM, Frank Setzpfandt wrote:
> I try to calculate the spatial dispersion relation (longitudinal k  
> as a function
> of transverse k) of a 1D-waveguide array at a certain wavelength.  
> Therefore I
> use the find-k routine of MPB. My .ctl file works well for the first  
> band but
> already takes a lot of time for the computation.
> To calculate also higher bands I would like to parallelize the  
> calculations but
> I'm not sure which is the best way to do so. Most of the ways (MPI,  
> k-list
> splitting) described in the wiki seem to be designed for the usual  
> run-routines
> to calculate w(k) and I'm not sure how to implement them with find-k.

If you use the MPI version of MPB, find-k will be parallelized without  
modification.