[gmx-users] Regarding the change in number of processors of simulation job

[Sujit Sarkar]

Hello everyone,

When I am restarting a simulation job after crash or extending simulation,
if I increase number of processor then sharp change in coulomb potential
(short range) is observed.
But, if I keep the number of processor fixed then such thing is not found.

My question is why this is observed? Secondly, do I have to always keep the
number of processor fixed throughout the whole simulation?