Khanyile, NPTapamo, J-RDube, E2012-09-172012-09-172012-08Khanyile, NP, Tapamo, J-R and Dube, E. 2012. An analytic model for predicting the performance of distributed applications on multicore clusters. IAENG International Journal of Computer Science, vol. 39(3), pp 312-3201819-656Xhttp://www.iaeng.org/IJCS/issues_v39/issue_3/IJCS_39_3_11.pdfhttp://hdl.handle.net/10204/6099IAENG International Journal of Computer Science, vol. 39(3), pp 312-320Computationally demanding applications can benefit from distributed processing. Distributed processing offers immerse computational power from collections of autonomous systems which can provide up to thousands of processing cores. Over the years, a considerable amount of research has been put towards developing performance prediction models for distributed applications. Amdahl's law states that the speedup of any parallel program has an upper bound which is determined by the amount of time spent on the sequential fraction of the program, no matter how small and regardless of the number of processing nodes used. This paper discusses some of the short comings of this law in the current age. We propose a theoretical model for predicting the behavior of a distributed algorithm given the network restrictions of the cluster used. The paper focuses on the impact of latency and bandwidth which affect the cost of interprocessor communication and the number of processing nodes used to predict the performance. The model shows good accuracy in comparison to Amdahl's law.enMulticore clustersAmdahl's lawAlgorithmsPropagation delayBandwithLatencyDistributed programmingArticleKhanyile, N., Tapamo, J., & Dube, E. (2012). An analytic model for predicting the performance of distributed applications on multicore clusters. http://hdl.handle.net/10204/6099Khanyile, NP, J-R Tapamo, and E Dube "An analytic model for predicting the performance of distributed applications on multicore clusters." (2012) http://hdl.handle.net/10204/6099Khanyile N, Tapamo J, Dube E. An analytic model for predicting the performance of distributed applications on multicore clusters. 2012; http://hdl.handle.net/10204/6099.TY - Article AU - Khanyile, NP AU - Tapamo, J-R AU - Dube, E AB - Computationally demanding applications can benefit from distributed processing. Distributed processing offers immerse computational power from collections of autonomous systems which can provide up to thousands of processing cores. Over the years, a considerable amount of research has been put towards developing performance prediction models for distributed applications. Amdahl's law states that the speedup of any parallel program has an upper bound which is determined by the amount of time spent on the sequential fraction of the program, no matter how small and regardless of the number of processing nodes used. This paper discusses some of the short comings of this law in the current age. We propose a theoretical model for predicting the behavior of a distributed algorithm given the network restrictions of the cluster used. The paper focuses on the impact of latency and bandwidth which affect the cost of interprocessor communication and the number of processing nodes used to predict the performance. The model shows good accuracy in comparison to Amdahl's law. DA - 2012-08 DB - ResearchSpace DP - CSIR KW - Multicore clusters KW - Amdahl's law KW - Algorithms KW - Propagation delay KW - Bandwith KW - Latency KW - Distributed programming LK - https://researchspace.csir.co.za PY - 2012 SM - 1819-656X T1 - An analytic model for predicting the performance of distributed applications on multicore clusters TI - An analytic model for predicting the performance of distributed applications on multicore clusters UR - http://hdl.handle.net/10204/6099 ER -