Nishant Agrawal, Abhishek Das, Girishchandra R. Yendargaye, T. M. Prabhu, Sandeep K. Joshi, V. V. Shenoi
{"title":"Performance Analysis of Python-based Finite Volume Solver ANUGA on Modern Architectures","authors":"Nishant Agrawal, Abhishek Das, Girishchandra R. Yendargaye, T. M. Prabhu, Sandeep K. Joshi, V. V. Shenoi","doi":"10.1145/3474124.3474152","DOIUrl":null,"url":null,"abstract":"The performance analysis of ANUGA, a Python-based finite volume solver on the unstructured grid for shallow water model in two dimensions, on recent Intel and AMD processor-based HPC clusters, form the focus of the present study. The analysis uses three datasets with different resolutions of the underlying triangular mesh for discretizing the region of interest. The computational workload depends on the resolution of the grid, which impacts the computation time required for the simulation of the flow. The factors influencing the parallel performance: workload distribution across different processes, and the bulk of the data exchange (the communication overhead), are studied systematically. This paper is an account of the preliminary study to understand the impact of the memory hierarchy (due to the relative sizes of cache) available on different architecture on the performance of this application.","PeriodicalId":144611,"journal":{"name":"2021 Thirteenth International Conference on Contemporary Computing (IC3-2021)","volume":"55 6","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2021-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2021 Thirteenth International Conference on Contemporary Computing (IC3-2021)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3474124.3474152","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
The performance analysis of ANUGA, a Python-based finite volume solver on the unstructured grid for shallow water model in two dimensions, on recent Intel and AMD processor-based HPC clusters, form the focus of the present study. The analysis uses three datasets with different resolutions of the underlying triangular mesh for discretizing the region of interest. The computational workload depends on the resolution of the grid, which impacts the computation time required for the simulation of the flow. The factors influencing the parallel performance: workload distribution across different processes, and the bulk of the data exchange (the communication overhead), are studied systematically. This paper is an account of the preliminary study to understand the impact of the memory hierarchy (due to the relative sizes of cache) available on different architecture on the performance of this application.