Title :
A Generic and Highly Efficient Parallel Variant of Boruvka´s Algorithm
Author :
Da Silva Sousa, Cristiano ; Mariano, Artur ; Proenca, Alberto
Abstract :
This paper presents (i) a parallel, platform independent variant of Boruvka´s algorithm, an efficient Minimum Spanning Tree (MST) solver, and (ii) a comprehensive comparison of MST-solver implementations, both on multi-core CPU-chips and GPUs. The core of our variant is an effective and explicit contraction of the graph. Our multi-core CPU implementation scales linearly up to 8 threads, whereas the GPU implementation performs considerably better than the optimal number of threads running on the CPU. We also show that our implementations outperform all other parallel MST-solver implementations in (ii), for a broad set of publicly available road network graphs.
Keywords :
microprocessor chips; multiprocessing systems; trees (mathematics); Boruvka´s Algorithm; GPU; minimum spanning tree solver; multicore CPU-chips; road network graphs; Algorithm design and analysis; Arrays; Color; Graphics processing units; Instruction sets; Kernel; boruvka; gpu; minimum spanning tree; parallel; performance;
Conference_Titel :
Parallel, Distributed and Network-Based Processing (PDP), 2015 23rd Euromicro International Conference on
Conference_Location :
Turku
DOI :
10.1109/PDP.2015.72