How to utilize x number of cores?

ARCA's Fortran model is not parallelized. Everything is calculated in linear fashion, meaning that the result of one timestep is vital to the the next one, and this lends itself poorly to parallelization. For the most part this is not a problem since single runs do not take too long, especially if the optimized time step is used. In batches however, the situation changes, and you might have hours or days of computing ahead of you. Of course, the best option is to use a supercomputer where the individual tasks can be divided to tens or more cores and the whole batch goes through very fast. On a personal computer, divide the batch files in two to four (depending on the number of cores, leave some for the system too) separate batches and run them all at the same time. Then you will really put your computer to work, but our experience shows roughly 2.5 times faster overall result (with an average laptop).