Given a sequence of chunk Bi xmi , where the xij denote characters of an alphabet Σ, we wish to apply the classical hash function h(Bi) = Bi mod P for some large prime number P. We assume the availability of several processors working in parallel. A simplistic solution of assigning a set of processors would be as follows. Suppose an (unbounded) stream of non-overlapping chunks B is given, we shall process them by subsets of elements. For each subset, the processors are assigned sequentially to the chunks. Once all of them have produced their output, the whole set of processors is reassigned to the following chunks, etc. We may assume that all the processors need roughly the same time for the processing of their respective chunks, since the execution time of the given operation, B mod P in our example, is not data dependent. Therefore this way of processing the stream by subsets of chunks keeps all the processors busy all of the time. We call this the basic parallel method.