is easy: each dataset can be run separately. Make different directories, ideally, and link the output (GS WFK in particular) into each as required, and then you can run the dtsets separately: GS first, then any phonon except Gamma, and Gamma requires the DDK calculation first, and is a first step before doing the Raman 3rd order DFPT. Optimally you can reduce it down to a single perturbation in each run
Actually, this is much simpler with abipy, and there is a workflow which does all this for you automatically!
See here: Phonons with DFPT β abipy 0.9.8 documentation
If DFPT see above. If you are using the optics driver I think it can run in parallel over k, perhaps bands as well, but I am not an expert there. Try with something small (4x4x4 k 10 bands) and look at your cpu usage, to make sure itβs doing what you expect.
Thanks for the response. In case 1, there are three DDK Wavefunctions are generated as well. How would I feed multiple wavefunctions i.e. outputs in my next step calculation? Besides, for my convenience I am not using the abipy codes, could you please tell me in terms of tutorials point of view?
make each pert in another directory, call it dtset 3 in each case, and symbolic link these files to the following Gamma point dtsets which need the ddk