HPC and ReaxFF

Dear Axel
Thanks for your Detailed explanations.
Ì know that the using ReaxFF, need more compotatonal cost in compare of conventional potential and I studied several literature about compartion of thease potentials. I am about 3 years on MD simulation.
I think we need to know what goal we pursue in our simulation. I have reached my results in small simulation boxes, but I want to know the effect of simulation box size on the result. Also I am very curious to know about the effect parameters in the cost calculations of using reaxFF but I am a beginner in this regard.