TU Darmstadt / ULB / TUbiblio

Optimizing the Memory Access Performance of FASTEST's Sispol Routine

Burger, Michael and Bischof, Christian (2014):
Optimizing the Memory Access Performance of FASTEST's Sispol Routine.
In: World Congress on Computational Mechanics, Barcelona, Spain, [Online-Edition: http://www.wccm-eccm-ecfd2014.org/admin/files/filePaper/p153...],
[Conference or Workshop Item]

Abstract

In this article the runtime behavior of the simulation software Fastest is investigated and its performance is optimized. The main performance bottleneck of Fastest is the sipsol subroutine. An analysis shows that the memory accessing behavior is the main reason for the high runtime. As a consequence, a new data structure is developed which on the one hand increases the speed but on the other hand preserves the ability to parallelize the calculations on the data grid executed within the sipsol subroutine. The ratio between calculation and memory reads/writes is improved significantly, and the performance in certain cases doubled. Finally, approaches to efficiently implement a parallelization within the new structure are proposed.

Item Type: Conference or Workshop Item
Erschienen: 2014
Creators: Burger, Michael and Bischof, Christian
Title: Optimizing the Memory Access Performance of FASTEST's Sispol Routine
Language: English
Abstract:

In this article the runtime behavior of the simulation software Fastest is investigated and its performance is optimized. The main performance bottleneck of Fastest is the sipsol subroutine. An analysis shows that the memory accessing behavior is the main reason for the high runtime. As a consequence, a new data structure is developed which on the one hand increases the speed but on the other hand preserves the ability to parallelize the calculations on the data grid executed within the sipsol subroutine. The ratio between calculation and memory reads/writes is improved significantly, and the performance in certain cases doubled. Finally, approaches to efficiently implement a parallelization within the new structure are proposed.

Volume: 11
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Scientific Computing
Exzellenzinitiative
Zentrale Einrichtungen
Zentrale Einrichtungen > University IT-Service and Computing Centre (HRZ) > Hochleistungsrechner
Zentrale Einrichtungen > University IT-Service and Computing Centre (HRZ)
Exzellenzinitiative > Graduate Schools > Graduate School of Computational Engineering (CE)
Exzellenzinitiative > Graduate Schools
Event Title: World Congress on Computational Mechanics
Event Location: Barcelona, Spain
Date Deposited: 26 Mar 2015 12:32
Official URL: http://www.wccm-eccm-ecfd2014.org/admin/files/filePaper/p153...
Alternative keywords:
Alternative keywordsLanguage
Data Structures, Performance Optimization, Parallelization, Cache OptimizationEnglish
Export:

Optionen (nur für Redakteure)

View Item View Item