GPU-BLAST [18] and CUDA-BLASTP [19] mainly exploit coarse-grained parallelism, i.e., one sequence alignment is mapped to only one thread. In ...
確定! 回上一頁