Research Article
Decision Scheduling for Cloud Computing Tasks Relying on Solving Large Linear Systems of Equations
Table 2
Experimental data of Gauss elimination method.
| Number of parallel nodes | Dense matrix scale | 1002 | 2002 | 5002 | 6002 | 9002 |
| | Total time | 0.053 | 0.0817 | 0.0093 | 0.1104 | 0.0523 | | Communication time | 0.0576 | 0.0994 | 0.0935 | 0.099 | 0.0254 | 1 | Parallel computing time | 0.0446 | 0.0064 | 0.0344 | 0.1106 | 0.0059 | | Speedup ratio | — | — | — | — | — | | Parallel efficiency | — | — | — | — | — | | Total time | 0.1128 | 0.0884 | 0.0534 | 0.0554 | 0.0488 | | Communication time | 0.0777 | 0.0155 | 0.0955 | 0.1121 | 0.0932 | 2 | Parallel computing time | 0.0574 | 0.0069 | 0.0928 | 0.1062 | 0.0052 | | Speedup ratio | 0.67 | 0.53 | 0.31 | 0.85 | 0.85 | | Parallel efficiency | 0.33 | 0.92 | 0.99 | 0.39 | 0.49 | | Total time | 0.1069 | 0.0932 | 0.0617 | 0.0501 | 0.0181 | | Communication time | 0.0988 | 0.0045 | 0.0873 | 0.0056 | 0.044 | 4 | Parallel computing time | 0.0101 | 0.0868 | 0.0439 | 0.0328 | 0.0838 | | Speedup ratio | 1.12 | 0.9 | 0.88 | 0.98 | 0.68 | | Parallel efficiency | 0.4 | 1.19 | 0.51 | 0.3 | 0.64 |
|
|