Abstract:
Improving power consumption and performance are the important
goals of scheduling on multiprocessors. In power
aware scheduling with dynamic voltage/frequency scaling
(DVFS), tasks are made to run on low voltages, which decreases
their computation power. However, it also increases
their execution costs and hence, may increase the schedule
length. Furthermore, applying DVFS on processors does
not impact the communication delay and power consumption.
Duplicating a task on multiple processors reduces the
communication delay among processors, which further reduces
the schedule length and improves the performance1
.
Additionally, duplication reduces the communication energy
among processors, but also increases the overall computation
energy. In this paper, we propose an integrated DVFS and
duplication based solution to schedule task graphs on heterogeneous
multiprocessors. The use of both techniques is optimized
with Mixed Integer Programming (MIP) formulation
to achieve better power and performance at the same time.
To enhance the MIP convergence, each task is run by integrating
the maximum and minimum voltage on a processor
instead of iterating through all the voltage levels. The results
demonstrate a minimum of 50% improvement in the processor
power and 20−50% improvement in the total power (processor
and communication) with a performance comparable
to the other algorithms.
