Research Article
An Efficient Evolutionary Task Scheduling/Binding Framework for Reconfigurable Systems
Table 1
Benchmark specifications.
| ID | Name | # of nodes | # of edges | Avg. edges/node | Critical path | Parallelism (nodes/crit. path) |
| S1 | Synthesized (4 task types) | 50 | 50 | 1 | 6 | 8.3 | S2 | Synthesized (4 task types) | 150 | 200 | 1.33 | 15 | 10.1 | S3 | Synthesized (8 task types) | 25 | 40 | 1.6 | 7 | 3.5 | S4 | Synthesized (8 task types) | 100 | 120 | 1.2 | 7 | 14.2 | S5 | Synthesized (5 task types) | 13 | 10 | 0.77 | 3 | 4.3 | S6 | Synthesized (4 task types) | 10 | 5 | 0.5 | 3 | 3.3 | S7 | Synthesized (8 task types) | 50 | 60 | 1.2 | 8 | 6.3 | S8 | Synthesized (4 task types) | 100 | 120 | 1.2 | 7 | 14.3 | S9 | Synthesized (8 task types) | 150 | 200 | 1.33 | 8 | 25.1 | S10 | Synthesized (8 task types) | 200 | 60 | 0.3 | 4 | 50.1 | S11 | Synthesized (4 task types) | 100 | 150 | 1.5 | 8 | 12.5 |
| DFG2 | JPEG-Smooth Downsample | 51 | 52 | 1.02 | 16 | 3.2 | DFG6 | MPEG-Motion Vectors | 32 | 29 | 0.91 | 6 | 5.3 | DFG7 | EPIC-Collapse_pyr | 56 | 73 | 1.3 | 7 | 8.1 | DFG12 | MESA-Matrix Multiplication | 109 | 116 | 1.06 | 9 | 12.1 | DFG14 | HAL | 11 | 8 | 0.72 | 4 | 2.8 | DFG16 | Finite Input Response Filter 2 | 40 | 39 | 0.975 | 11 | 3.6 | DFG19 | Cosine 1 | 66 | 76 | 1.15 | 8 | 8.3 |
|
|