An Improved Smith-Waterman Algorithm on Heterogeneous CPU-GPU Systems

Smith-Waterman (SW) algorithm is a classic dynamic programming algorithm to solve the problem of biological sequence alignment. This paper describes the approach and the speedup obtained in performing Smith-Waterman database searches on heterogeneous platforms comprising of CPU and GPU systems. The improved parallelization Smith-Waterman algorithms based on column, which is rather than Liu’s algorithm which computes all the elements in the same anti-diagonal independently of each other in parallel.In order to enhance the parallelism, we exploit both the task-level parallelism and the data-level parallelism.Some optimization strategies are used in our implementation, we circlularly use the shared memory, compute the maximum used the method similar to reduction, realized with the coalesced access the global memory. The experiment result shows that our new parallel algorithm is more efficient than that of previous.