Table 5 CS-4 running times (milliseconds) of MPI implementation
From: Solving the maximum subsequence sum and related problems using BSP/CGM model and multi-GPU CUDA
n | N16:P1 | N16:P2 | N16:P4 | N16:P8 | N32:P1 | N32:P2 | N32:P4 | N32:P8 |
|---|---|---|---|---|---|---|---|---|
220 | 10.730 | 11.403 | 9.788 | 91.368 | 11.597 | 19.765 | 21.263 | 119.741 |
221 | 21.403 | 15.286 | 12.880 | 83.876 | 16.562 | 21.004 | 18.488 | 79.335 |
222 | 99.957 | 23.938 | 18.761 | 16.064 | 24.395 | 25.461 | 21.225 | 23.279 |
223 | 82.786 | 40.277 | 29.675 | 21.406 | 45.941 | 33.450 | 26.846 | 28.399 |
224 | 175.851 | 75.340 | 51.073 | 33.305 | 82.706 | 52.493 | 38.760 | 33.242 |
225 | 357.882 | 139.970 | 90.844 | 53.800 | 146.140 | 93.536 | 59.839 | 46.379 |
226 | 748.831 | 263.430 | 161.765 | 100.714 | 263.629 | 164.268 | 93.531 | 66.719 |
227 | 1546.020 | 559.685 | 307.145 | 171.782 | 555.407 | 288.917 | 170.073 | 111.906 |
228 | – | 1331.303 | 2356.562 | 314.681 | 854.059 | 426.578 | 317.116 | 203.811 |
229 | – | – | – | 24,301.072 | – | 4174.692 | – | 17,766.883 |