Graphics processing units (GPUs) are increasingly applied to accelerate tasks such as graph problems and discreteevent simulation that are characterized by irregularity, i.e., a strong dependence of the control flow and memory accesses on the input. The core data structure in many of these irregular tasks are priority queues that guide the progress of the computations and which can easily become the bottleneck of an application. To our knowledge, currently no systematic comparison of priority queue implementations on GPUs exists in the literature. We close this gap by a performance evaluation of GPU-based priority queue implementations for two applications: discrete-event simulation and parallel A* path searches on grids. We focus on scenarios requiring large numbers of priority queues holding up to a few thousand items each. We present performance measurements covering linear queue designs, implicit binary heaps, splay trees, and a GPU-specific proposal from the literature. The measurement results show that up to about 500 items per queue, circular buffers frequently outperform tree-based queues for the considered applications, particularly under a simple parallelization of individual item enqueue operations. We analyze profiling metrics to explore classical queue designs in light of the importance of high hardware utilization as well as homogeneous computations and memory accesses across GPU threads.

In the context of a master level programming practical at the computer science department of the Karlsruhe Institute of Technology, we developed and make available two independent and highly optimized open-source implementations for the pair-wise statistical alignment model, also known as TKF91, that was developed by Thorne, Kishino, and Felsenstein in 1991. This paper has two parts. In the educational part, we cover teaching issues regarding the setup of the course and the practical and summarize student and teacher experiences. In the scientific part, the two student teams (Team I: Nikolai, Sebastian, Daniel; Team II: Sarah, Pierre) present their solutions for implementing efficient and numerically stable implementations of the TKF91 algorithm. The two teams worked independently on implementing the same algorithm. Hence, since the implementations yield identical results -with slight numerical deviations-we are confident that the implementations are correct. We describe the optimizations applied and make them available as open-source codes in the hope that our findings and software will be useful to the community as well as for similar programming practicals at other universities.