Title: Performance Evaluation of Matrix Multiplication on a Network of Work Stations with Communication Delay
Abstract: Network-based multicomputer systems merge as a potential economical candidate to replace supercomputers. Despite enormous effort to evaluate the performance of those systems, we still lack a complete performance model to analyze distributed computing systems. The model is complete if all system parameters, network parameters, communication overhead parameters, and application parameters are considered explicitly in the solution. In this paper we develop a closed form solution for a finish time of executing a matrix-multiplication on a distributed network of processors (cloud computing). Matrix multiplication represents a paradigm of divisible jobs, which has a wide area of applications. The closed form solution is derived considering one-dimensional decomposition and two-dimensional decomposition of the matrices involved in the operation. The analytical results combine all system parameters, network parameters, and matrix dimension in the solution, which helps the designer to study the effect of each individual parameter on the overall system performance. This then becomes a tool for a designer of a multicomputer system to mange limited resources in an optimal manner paying attention only to those parameters that are most critical. The results show the superiority of the row-wise one-dimensional decomposition. We also show that, because of the communication overhead, adding more processors will not incessantly improve the system performance. We derived a closed form solution for the optimum number of processors (nodes) to be used to obtain the maximum speed up (equally the minimum finish time). Process allocation tradeoff is very important for optimizing the multicomputer system performance.
Publication Year: 2011
Publication Date: 2011-01-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 2
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot