The potential for parallel processing and neural computing in management decision support

Abstract

Some of the current parallel-processing architectures are discussed, and management-science-type problems amenable to solution using coarse-grained multiple-instruction, multiple-data parallel computers are identified. Applications of neural computing in management science are also discussed. It is shown that the real promise of parallelism is not just in doing the same things faster, but in routinely solving problems which were previously extraordinarily difficult or even impossible to solve, or where the solution was not economically justifiable because of the large computational effort. In a qualitative sense, parallel processing enables approaching certain problems from an entirely different perspective.<>