Performance Optimization of Aho-Corasick Algorithm on a GPU

Aho-Corasick (AC) algorithm is a multiple patterns matching algorithm commonly used for applications such as computer and network security, bioinformatics, image processing, among others. These applications are computationally demanding, thus optimizing performance for AC algorithm is crucial. In this paper, we present a performance optimization strategy for the AC algorithm on a Graphic Processing Unit (GPU). Our strategy efficiently utilizes the high degree of the on-chip parallelism and the complicated memory hierarchy of the GPU so that the aggregate performance (or throughput) for the AC algorithm can be optimized. The strategy significantly cuts down the effective memory access latencies and efficiently utilizes the memory bandwidth. Also, it maximizes the effects of the multithreading capability of the GPU through optimal thread scheduling. Experimental results on Nvidia GeForce GTX 285 GPU show that our approach delivers up to 127 Gbps throughput performance and 222-times speedup compared with a serial version running on single core of 2.2Ghz Core2Duo Intel processor.