Similar to divide-and-conquer, but avoids duplicate work when subproblems are identical.
(Typically used for optimization problems like the Traveling Salesman Problem).