Numerical Analysis,
Computer Science
A class of
mathematical problems that are extremely amenable to
parallel processing. Many papers have been written on special purpose parallel algorithms to speed up the
wallclock time of a given algorithm. These algorithms are invariably constrained by inherent seriality of the problem. Certain kinds of problems have such easy parallelization that one is almost "embarrassed" to talk about how simple it is to get many processors to be used efficiently. One example is
genetic algorithms, in which multiple generations of simulations can be operated simultaneously.