Publication Details
Daming Feng, Andrey Chernikov and Nikos Chrisochoides.
Published in Parallel Computing, 10.1016/j.parco.2016.01.007, Volume 59, pages 60 -- 70, November, 2016
Abstract
This paper presents a three dimensional two-level Locality- Aware Parallel Delaunay image-to-mesh conversion algorithm (LAPD). The algorithm exploits two levels of parallelism at different granularities: coarse-grain parallelism at the region level (which is mapped to a node with multiple cores) and medium-grain parallelism at the cavity level (which is mapped to a single core). LAPD employs a data locality-aware mesh re cement process to reduce the latency caused by the remote memory access. The LAPD method is evaluated on Blacklight, a cache-coherent NUMA distributed shared memory (DSM) machine in the Pittsburgh Supercomputing Center, and observed a weak scaling efficiency of almost 70% for roughly 200 cores, compared to only 30% for the previous algorithm, Parallel Optimistic Mesh Generation algorithm (PODM). To the best of our knowledge, LAPD exhibits the best scalability for parallel Delaunay mesh generation algorithms running on NUMA DSM supercomputers.
