I generate a binary output file using write_restart comment.
Later I used same binary file to simulate two similar runs (same operations and same number of steps).
But the output (dump file- (system volume)) from two cases is different. Why is this so?
I simulate same operation, but used different number of processors (parallelization) 4 and 24.
These are two cases i obtained at 1 atm pressure and 83 K.
will the number of processor makes such changes in result?
Impossible to say without knowing more details (LAMMPS version, how compiled, what platform, what exact command line used, etc.) and providing the exact input decks so one can try to reproduce this behavior. It would also be useful to know, how many times you have tried this run and whether you had made this experiment multiple times with additional number of processors or different types of parallelization.
In general, the explanations given here (11.1. Common problems — LAMMPS documentation) apply, but there is also a chance, that you may be running into a bug in the code. The latter cannot be resolved without having the input decks and being able to reproduce the unexpected behavior independently.
It is also possible, that you did not use the exact same input, but forgot or didn’t notice.