On the parallel and vlsi implementation of the interior point algorithms

摘要

ew discuss several aspects of the parallel and the VLSI implementa- tion of the interior point algorithms for solving the linear program- ming problem. Various architectures, based on the different parti- tioning of the input data, are proposed and the complexity of these implementations are discussed. Am economic interpretation of a par- titioning scheme is then presented which might prove useful for the further improvements in the efficient parallelization of the interior point methods.