www.neuron.yale.edu

My simulations have almost the same configurations (total number of cells, how the cells are connected etc.) Except, the properties of interest, for example , different input frequency to the neural network.
I've cheated on parallel simulations by running multiple 'hoc' files with no parallel programming code on a single machine.
(I'll move to parallel programming, of course, when there are more neurons on the network that the single simulation is too slow to handle.)

So far, the results are ok. The same configuration will give the same results when run by different NEURON instance at the same time.
However, I know that this method is just a crude method and there might be some system conflicts that I have not aware of.
I wonder what are those potential errors that could happen when I do it this way(if any). Or the different NEURON instances will always work independently from each other and I could continue working this way.

Thank you so much.

There shouldn't be any hidden traps.

Just to confirm, that's mean I can safely continue doing it, right? ( as long as my single simulation is not too unmanageable slow)

The simulations won't interfere with each other, unless they try to write to the same output file or consume so much memory that execution is slowed. If either of those outcomes happens, you'll recognize that something is wrong.