Data-parallel halo finding with variable linking lengths

摘要: State-of-the-art cosmological simulations regularly contain billions of particles, providing scientists the opportunity to study evolution Universe in great detail. However, rate at which these generate data severely taxes existing analysis techniques. Therefore, developing new scalable alternatives is essential for continued scientific progress. Here, we present a dataparallel, friends-of-friends halo finding algorithm that provides unprecedented flexibility by extracting multiple linking lengths. Even single length, it as fast techniques, and portable multi-threaded many-core systems well co-processing resources. Our system implemented using PISTON coupled an interactive environment used halos different lengths track their over time.