i am wondering at this problem. i was told that, if the timestep is too small, then the simulation runs very slow,if it is too big, the the error will increase. so we should choose a right timestep with acceptable error and maximum speed. but what is the criteria? i have looked at some inp file of charmm, and found they mostly choose a timestep of 0.002 ps. but i do not know why?
thank you for your kind help.