The SA algorithm does not require or deduce derivative information, it merely needs to be supplied with an objective function for each trial solution it generates. Thus, the evaluation of the problem functions is essentially a `black box' operation as far as the optimization algorithm is concerned. Obviously, in the interests of overall computational efficiency, it is important that the problem function evaluations should be performed efficiently, especially as in many applications these function evaluations are overwhelming the most computationally intensive activity. Depending on the nature of the system equations advice on accelerating these calculations may be found in other chapters within this project.

Some thought needs to be given to the handling of constraints when using the SA algorithm. In many cases the routine can simply be programmed to reject any proposed changes which result in constraint violation, so that a search of feasible space only is executed. However, there are two important circumstances in which this approach cannot be followed:

- if there are any equality constraints defined on the system, or
- if the feasible space defined by the constraints is (suspected to be) disjoint, so that it is not possible to move between all feasible solutions without passing through infeasible space.