Anomalies in parallel branch-and-bound algorithms

<italic>We consider the effects of parallelizing branch-and-bound algorithms by expanding several live nodes simultaneously. It is shown that it is quite possible for a parallel branch-and-bound algorithm using n</italic><subscrpt>2</subscrpt> <italic>processors to take more time than one using n</italic><subscrpt>1</subscrpt> <italic>processors, even though n</italic><subscrpt>1</subscrpt> < <italic>n</italic><subscrpt>2</subscrpt>. <italic>Furthermore, it is also possible to achieve speed-ups that are in excess of the ratio n</italic><subscrpt>2</subscrpt>/<italic>n</italic><subscrpt>1</subscrpt>. <italic>Experimental results with the 0</italic>/<italic>1-Knapsack and Traveling Salesman problems are also presented.</italic>