Análise de sensibilidade da variação quantitativa de métricas de qualidade sobre a eficiência de um Raytracer

Abstract

Currently, large-scale manufacture of multiprocessor systems has been accompanied by a review of techniques for parallel software development. However, it is clear that the way in which parallel applications are being developed is oriented primarily to reduce the execution time, being disregarded, thus traditional metrics of software quality such reusability, maintainability and scalability. In addition, other metrics related to efficiency such energy consumption and memory, especially in important emerging areas in the field of software development such as embedded systems, has also been neglected. Therefore, the objective of this study is to demonstrate a relationship between software quality metrics, in favor of software reuse, maintainability and reducing the execution time and efficiency metrics such as cycles, energy, memory consumption, among others, thereby potentially identify impacts that the change in software metrics have on the efficiency of a parallel application when run on an embedded multiprocessor platform.