Optimisation of Mapping and Scheduling for Dynamic Voltage Scaling
This section has analysed the effect of application mapping under different constellations of scheduling and voltage scaling techniques. The conducted experiments have shown that substantial energy saving can be achieved by the iterative techniques introduced in this book,when compared to constructive scheduling and mapping approaches [138, 158 ], which have been used in previous work on energy minimisation through DVS [20, 99]. Furthermore, the experiments indicate an advantage of the proposed techniques over constructive approaches also in terms of schedulability in the presence of tasks with tight deadlines. The results reinforce the importance of a thorough exploration of the mapping and scheduling solution space. Clearly, the cost for these better results is higher computational time.
KeywordsPower Dissipation Task Mapping Communication Mapping Task Graph List Schedule
Unable to display preview. Download preview PDF.
- 1.For distributed systems consisting of several processing elements and communication links, a ready list is introduced for each component. Furthermore, priorities are additionally assigned to communications.Google Scholar
- 2.Energy-Efficient Genetic List Scheduling AlgorithmGoogle Scholar
- 3.The DLSP results are identical to the results presented in Table 3.5 (Section 3.3.1).Google Scholar