Revisiting Serial Arithmetic: A Performance and Tradeoff Analysis for Parallel Applications on Modern FPGAs

Author(s):  
Aaron Landy ◽  
Greg Stitt
2019 ◽  
Vol 26 (2) ◽  
pp. 45-54
Author(s):  
Gabriella Lopes Andrade ◽  
Marcia Cristina Cera

Run tracing allows you to identify issues affecting the performance of parallel applications. This work consists in evaluating the parallelization of a Genetic Algorithm applied to the Vehicle Routing Problem with OpenMP, where the performance obtained was not ideally expected. Being that it was possible to obtain a performance increase of 1.4 times in the architecture used, however, but still below ideal. Therefore, the general objective of this work is to investigate the causes of the low performance obtained by the Genetic Algorithm, performing an analysis from the execution traces. Our results showed that the parallelization of the Genetic Algorithm is according to the model in which it was implemented and to the set of instances of the target Vehicle Routing Problem used.


2016 ◽  
Vol 65 (2) ◽  
pp. 495-507 ◽  
Author(s):  
Weizhe Zhang ◽  
Albert M. K. Cheng ◽  
Jaspal Subhlok

Sign in / Sign up

Export Citation Format

Share Document