GPU-accelerated processing of polygonal meshes in compact space

Abstract

We present a new compact half-edge data structure for querying and traversing polygonal meshes on the GPU. For this, we adapted a planar graph data structure called PEMB to be used inside a GPU. To test, we built a version of the mesh generator Polylla in GPU compact space and compared it against its GPU non-compact, sequential non-compact, and sequential compact versions. The latter takes a triangulation and generates a polygonal mesh, while our algorithm outputs a compact mesh. We also measured and compared the performance of queries from the data structures. The results show that generating meshes using the GPU compact variant is 2x faster than the sequential compact variant. Moreover, queries from the data structure are 43x faster using the GPU compact variant compared to the sequential one.

Diego García Espinoza

Paralelización de estructura de datos compacta en GPU para la representación de figuras geométricas

Master Student of Computer Science at Universidad de Chile

Sergio Salinas-Fernández

Professor PEX at the Universidad de Chile

My research interests include Data science, Computational geometry and GPU computing.

José Fuentes-Sepúlveda

Associate Professor at University of Concepción

Associate Professor interested in the design and implementation of multicore algorithms, parallel construction of compact data structures and compression algorithms.

Nancy Hitschfeld Kahler

+Lab founder | Full Professor Universidad de Chile

Full Professor at the Department of Computer Science, University of Chile. Her main research interests include geometric modeling, geometric meshes, and parallel algorithms (GPU computing), focused in computational science, and engineering applications.