Plataforma web para medir el rendimiento de algoritmos en estructuras de datos. Apoyo al aprendizaje.
Loading...
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
Esta memoria presenta el desarrollo de Performance System, una plataforma web orientada al análisis empírico del rendimiento y consumo energético de algoritmos en C y C++. El sistema permite a estudiantes de la asignatura Estructuras de Datos ejecutar implementaciones propias en un entorno controlado y visualizar métricas categorizadas en CPU, memoria, sistema, tiempo y energía.
La propuesta consolida versiones anteriores de la plataforma mediante una arquitectura cliente–servidor con nodos de medición (slaves), incorporando validaciones de seguridad, límites de tiempo y registros administrativos. Se implementaron mejoras pedagógicas como un tutorial interactivo, ejemplos guiados y gráficos con barras de error que facilitan la interpretación de resultados.
El sistema fue desplegado en la infraestructura institucional de la Universidad de Concepción, quedando disponible en línea para su uso docente. Los experimentos realizados con algoritmos clásicos —Fibonacci, Longest Common Substring y ordenamientos por comparación evidencian que la plataforma revela de manera clara la relación entre complejidad algorítmica, patrones de acceso a memoria y consumo energético.
En síntesis, Performance System constituye un recurso didáctico robusto y accesible que integra teoría y evidencia empírica, contribuyendo a la formación en eficiencia computacional y sostenibilidad tecnológica.
This thesis presents the development of Performance System, a web-based platform designed to empirically analyze the performance and energy consumption of algorithms written in C and C++. The system enables students of the Data Structures course to execute their own implementations in a controlled environment and visualize categorized metrics in CPU, memory, system, time, and energy. The proposed solution consolidates previous versions of the platform through a client–server architecture with measurement nodes (slaves), integrating security validations, timeouts, and administrative logging. Pedagogical enhancements include an interactive tutorial, guided examples, and error-bar plots that support the interpretation of results. The system was deployed on the institutional infrastructure of the University of Concepción, making it available online for academic use. Experiments with classic algorithms—Fibonacci, Longest Common Substring, and comparison-based sorting—demonstrate that the platform effectively reveals the relationship between algorithmic complexity, memory access patterns, and energy consumption. In summary, Performance System constitutes a robust and accessible educational resource that bridges theoretical analysis with empirical evidence, contributing to training in computational efficiency and technological sustainability.
This thesis presents the development of Performance System, a web-based platform designed to empirically analyze the performance and energy consumption of algorithms written in C and C++. The system enables students of the Data Structures course to execute their own implementations in a controlled environment and visualize categorized metrics in CPU, memory, system, time, and energy. The proposed solution consolidates previous versions of the platform through a client–server architecture with measurement nodes (slaves), integrating security validations, timeouts, and administrative logging. Pedagogical enhancements include an interactive tutorial, guided examples, and error-bar plots that support the interpretation of results. The system was deployed on the institutional infrastructure of the University of Concepción, making it available online for academic use. Experiments with classic algorithms—Fibonacci, Longest Common Substring, and comparison-based sorting—demonstrate that the platform effectively reveals the relationship between algorithmic complexity, memory access patterns, and energy consumption. In summary, Performance System constitutes a robust and accessible educational resource that bridges theoretical analysis with empirical evidence, contributing to training in computational efficiency and technological sustainability.
Description
Tesis presentada para a optar al título de Ingeniero/a Civil Informático/a.
Keywords
Algoritmos computacionales, Algoritmos Procesamiento de datos, Procesamiento electrónico de datos