[Q-e-developers] Custom grids in hybrids

[Paolo Giannozzi]

While trying to port the reduced-cutoff machinery for computation of
(Vx\psi)(G) to the k-point case, I noticed that there are currently two
different FFT grids, one for the G=>r case, one for the r=>G case. I do not
understand the need for such distinction. Some tests show that merging the
two has no effect. Unless somebody gives a reason for having two different
FFTs, I'll proceed to merge the two (and to simplify the machinery, which
also looks to me complex beyond need)

Paolo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20150602/06841d57/attachment.html>