®®®® SIIA Público

Título del libro: Contemporary Topics In Computer Graphics And Games
Título del capítulo: Fast data parallel radix sort implementation in directx 11 compute shader to accelerate ray tracing algorithms

Autores UNAM:
ULISES OLIVARES PINTO;
Autores externos:

Idioma:

Año de publicación:
2019
Palabras clave:

Algorithms; Radix sort; Ray tracing


Resumen:

In this paper we present a fast data parallel implementation of the Radix Sort on the Direct Compute software development kit (SDK). We also discuss in detail the various optimization strategies that were used to increase the performance of our Radix Sort, and we show how these strategies can be generalized for any video card that supports the Direct Compute model. The insights that we share in this paper should be of use to any General-Purpose Graphics Processing Unit (GPGPU) programmer regardless of the video card being used. Finally, we discuss how Radix Sort can be used to accelerate ray tracing and we present some results in this area as well. © Peter Lang GmbH Internationaler Verlag der Wissenschaften Berlin 2020. All rights reserved.


Entidades citadas de la UNAM: