Skip to main content

Table 2 Performance Evaluation of CUDASW++

From: CUDASW++: optimizing Smith-Waterman sequence database searches for CUDA-enabled graphics processing units

Queries Length Single-GPU version Multi-GPU version
   Time(s) GCUPS Time(s) GCUPS
P02232 144 2.33 9.039 1.97 10.660
P01111 189 3.01 9.163 2.47 11.194
P05013 189 3.01 9.163 2.40 11.513
P14942 222 3.48 9.333 2.87 11.304
P00762 246 3.82 9.402 3.10 11.591
P07327 375 5.80 9.453 4.21 13.034
P01008 464 7.08 9.582 4.67 14.532
P10635 497 7.66 9.483 4.99 14.566
P25705 553 8.61 9.390 5.49 14.711
P03435 567 8.72 9.507 5.52 15.022
P42357 657 10.11 9.496 6.50 14.777
P21177 729 11.16 9.552 7.00 15.225
Q38941 850 13.02 9.539 8.15 15.235
O60341 852 13.03 9.556 8.11 15.355
P27895 1000 15.13 9.660 9.34 15.644
P07756 1500 22.74 9.642 14.02 15.640
P04775 2005 30.37 9.649 18.48 15.855
P19096 2504 37.89 9.659 22.79 16.058
P28167 3005 45.54 9.644 27.41 16.027
P0C6B8 3564 54.01 9.644 32.45 16.055
P20930 4061 61.60 9.635 36.94 16.070
P08519 4548 68.98 9.637 41.45 16.039
Q7TMA5 4743 71.91 9.640 43.18 16.054
P33450 5147 78.13 9.629 46.83 16.066
Q9UKN1 5478 83.15 9.629 49.77 16.087