Thank you for calling me the Japanese way.
Thank you for your interest and suggestion.
I am now using some workstations with AMD EPYC 7763 (64Core, 2.45GHz), with Intel Xeon 6154 (3.0GHz 18 Core ) and so on.
Your suggestion is fruitful.
Such a suggestion might be the greatest benefit of this community.
Certainly, although the optimizing process is iterative, the loops or sums within the iteration should be made faster by parallelization.