An Architecture for Improving the Efficiency of Specialized Vertical Search Engine Based on GPGPUs

In this paper, we propose a new architecture for improving the efficiency of specialized vertical search engine based on GPGPUs (General-purpose Graphics Processing Unit), so that the efficiency of information retrieval can be improved. The advantages of GPGPUs in programmability, high parallel processing ability, and floating-point computing power, etc. are utilized in the new architecture. Thus the CPU usage can be reduced and the efficiency of the search engine can be improved. Experiments show that a platform based on the proposed architecture has significantly higher efficiency.