We analyze the runtime behavior of an ant colony optimization approach for the longest common subsequence problem executed on a many-core GPU and a multi-core CPU. Our approach is a parallelized variant of a previously published algorithm. Moreover, we are able to significantly improve the results of the original one by adapting the heuristic function of the ant colony algorithm. Our results show that despite its many more cores the GPU has no significant advantages over the CPU-based approach.
CITATION STYLE
Markvica, D., Schauer, C., & Raidl, G. R. (2015). CPU versus GPU parallelization of an ant colony optimization for the longest common subsequence problem. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9520, pp. 401–408). Springer Verlag. https://doi.org/10.1007/978-3-319-27340-2_50
Mendeley helps you to discover research relevant for your work.