High Performance Biological Pairwise Sequence Alignment: FPGA versus GPU versus Cell BE versus GPP

<table>Our GPU parallel thread implementation of the Smith-Waterman algorithm: store and load operations are performed by the final thread and the first thread in each thread batch (block) to allow for any sequence length processing.</table>

International Journal of Reconfigurable Computing

fig9

Figure 9

Figure 9: High Performance Biological Pairwise Sequence Alignment: FPGA versus GPU versus Cell BE versus GPP 