In this paper, we show that we can achieve high performance in homology search by only adding one off-the-shelf PCI board with one Field Programmable Gate Array (FPGA) to a Pentium based computer system in use. By using off-the-shelf boards, we Carl easily obtain latest FPGAs at low costs which is very important because the performance is almost proportional to the vice of the FPGA, and FPGAs are becoming larger and larger following Moore's law. The performance can be furthermore accelerated by using more number of FPGA boards. In our approach, the search is divided into two phases, and different circuits are configured on the FPGA in each phase, its order to make up limited hardware resources of the off-the-shelf boards. The performance is almost comparable with small to middle class dedicated hardware systems when we use a board with one of the latest FPGAs. The computation time is almost proportional to the size of the query sequence. The time for comparing a query sequence of 2,048 elements with a database sequence of 64 millions elements by the Smith-Waterman algorithm is about 34 sec, which is about 330 times faster than a desktop computer with a 1GHz Pentium-III.
展开▼