This question was already discussed here, with a temporary solution by Mostapha.
So I´m just sharing my inputs on it.
For the new HB+ it´s possible to just write the files and then run them on different OS where you can benefit from more cores-so the simulation time will be much less.
But when you change OS, you run into some problems as backward and forward slashes in the path names etc. - therefore the attached script is useful.
In the future a more sophisticated way would be appreciated e.g. while writing the file, having a possibility to choose what is the platform on which it will be run + number of cores in case of Linux based OS.