Problem Decomposition in Parallel Networks
نویسندگان
چکیده
The improving computational speed of workstations makes clusters of these machines attractive for parallel computing. The workstation model of parallel processing, however, presents speciic challenges caused by the latency of the communications network and the workload imbalance that arises from the heterogeneity of the nodes. Because of these issues, data partitioning is critically important for parallel processing on workstation clusters. In this paper, we mathematically characterize the communication costs for four data decomposition schemes: scatter, contiguous point, contiguous row, and block. These methods are analyzed in terms of problem size, number of processors, network speed, communication pattern, and bookkeeping requirements. Bounds are established for the performance of these decomposition schemes that can be used to make better-informed data partitioning decisions.
منابع مشابه
A Benders' Decomposition Method to Solve Stochastic Distribution Network Design Problem with Two Echelons and Inter-Depot Transportation
In many practical distribution networks, managers face significant uncertainties in demand, local price of building facilities, transportation cost, and macro and microeconomic parameters. This paper addresses design of distribution networks in a supply chain system which optimizes the performance of distribution networks subject to required service level. This service level, which is consider...
متن کاملForward kinematic analysis of planar parallel robots using a neural network-based approach optimized by machine learning
The forward kinematic problem of parallel robots is always considered as a challenge in the field of parallel robots due to the obtained nonlinear system of equations. In this paper, the forward kinematic problem of planar parallel robots in their workspace is investigated using a neural network based approach. In order to increase the accuracy of this method, the workspace of the parallel robo...
متن کاملA Hybrid Solution Approach Based on Benders Decomposition and Meta-Heuristics to Solve Supply Chain Network Design Problem
Supply Chain Network Design (SCND) is a strategic supply chain management problem that determines its configuration. This mainly focuses on the facilities location, capacity sizing, technology selection, supplier selection, transportation, allocation of production and distribution facilities to the market, and so on. Although the optimal solution of the SCND problem leads to a significant reduc...
متن کاملSeries-Parallel and Tree-Decomposition Approaches for Fuzzy Constraint Networks
In this work, we present a Fuzzy Constraint Networks model for discrete domains, which generalizes the Disjunctive Fuzzy Temporal Constraint Networks model for temporal reasoning, and we propose the use of the seriesparallel and tree-decomposition approaches for simplifying its processing. After a separate empirical evaluation process of both techniques, a combined evaluation process over the s...
متن کاملStatic Task Allocation in Distributed Systems Using Parallel Genetic Algorithm
Over the past two decades, PC speeds have increased from a few instructions per second to several million instructions per second. The tremendous speed of today's networks as well as the increasing need for high-performance systems has made researchers interested in parallel and distributed computing. The rapid growth of distributed systems has led to a variety of problems. Task allocation is a...
متن کامل