Use this link to cite:
https://hdl.handle.net/2183/48870 Optimización y paralelización de un motor de físicas 2D para sistemas embebidos
Loading...
Identifiers
Publication date
Authors
Vázquez Cancela, Nicolás
Advisors
Other responsabilities
Universidade da Coruña. Facultade de Informática
Journal Title
Bibliographic citation
Type of academic work
Abstract
[Resumen]: Debido a la necesidad de interactuar en mundos virtuales con cuerpos físicos que responden de forma realista y en tiempo real, los videojuegos y otras experiencias interactivas han impulsado la investigación y desarrollo de los motores de físicas. Sin embargo, a causa de estos dos requisitos, y la complejidad de los cálculos que conlleva simular las colisiones entre objetos, siguen siendo uno de los procesos más costosos en la ejecución de videojuegos, sobre todo en tres dimensiones. Por estos motivos, se decidió mejorar el rendimiento del popular motor de físicas Box2D-lite, que por su limpia implementación, es la oportunidad perfecta de poner a prueba los conocimientos adquiridos en este máster para analizar e implementar soluciones óptimas, mediante técnicas sofisticadas de optimización y paralelización, que aceleraran notablemente
su velocidad de simulación. Además, para hacer más notables las restricciones del hardware y disponer de un mayor control de él a bajo nivel, se optó por adaptar y optimizar Box2D-lite a los sistemas embebidos Raspberry Pi.
[Abstract]: Due to the need to interact in virtual worlds with physical bodies that respond realistically and in real time, video games and other interactive experiences have driven the research and development of physics engines. However, because of these two requirements and the complexity of the calculations involved in simulating collisions between objects, they remain one of the most expensive processes in video game execution, especially in three dimensions. For these reasons, it was decided to improve the performance of the popular physics engine Box2D-lite, which, due to its clean implementation, provides the perfect opportunity to put the knowledge acquired in this master’s program to the test by analyzing and implementing optimal solutions through sophisticated optimization and parallelization techniques that significantly accelerate its simulation speed. Furthermore, to make hardware constraints more apparent and to achieve greater low-level control, we chose to adapt and optimize Box2D-lite for Raspberry Pi embedded systems.
[Abstract]: Due to the need to interact in virtual worlds with physical bodies that respond realistically and in real time, video games and other interactive experiences have driven the research and development of physics engines. However, because of these two requirements and the complexity of the calculations involved in simulating collisions between objects, they remain one of the most expensive processes in video game execution, especially in three dimensions. For these reasons, it was decided to improve the performance of the popular physics engine Box2D-lite, which, due to its clean implementation, provides the perfect opportunity to put the knowledge acquired in this master’s program to the test by analyzing and implementing optimal solutions through sophisticated optimization and parallelization techniques that significantly accelerate its simulation speed. Furthermore, to make hardware constraints more apparent and to achieve greater low-level control, we chose to adapt and optimize Box2D-lite for Raspberry Pi embedded systems.
Description
Editor version
Rights
Attribution 4.0 International







