• We need to concern ourselves with ensemble choice, equilibration and statistical niceties.

  • Samples are usually near the nanoscale and we may need quite a few to cover possible variations, especially for disordered systems.

  • Finite temperature simulations need special care.

  • There are good packages, but as with all computations care is also needed in their use.
  • Atomistic simulation is very easy what to do very badly what.
  • [BACK][NEXT]