A new scheduling technique based on Dynamic Voltage Scaling for MPSoC
Chang‐Woo Park, Kyung-Woo Noh, Seok-Yoon Kim
Abstract
Chang‐Woo Park, Kyung-Woo Noh, Seok-Yoon Kim
Abstract
The DVS(Dynamic Voltage Scaling) technique is the method to reduce the dynamic energy consumption. Using slack times, it extends the execution time of big load operations by changing frequency and voltage of variable voltage processors. Many researches have been going on to control the energy consumption of the processors and the data transmission among processors by controlling the bandwidth to reduce the energy consumption of the entire system. Since operations in multiprocessor systems have the data dependency between processors, however, the DVS techniques devised for single processors are not suitable to improve the energy efficiency of multiprocessor systems. We propose a new scheduling algorithm based on DVS for increasing energy efficiency of multiprocessor systems. The proposed DVS-based scheduling algorithm can improve the energy efficiency of entire systems because it controls frequency and voltages while considering the data dependency among processors.
OpenAlex reports 3 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The DVS(Dynamic Voltage Scaling) technique is the method to reduce the dynamic energy consumption. Using slack times, it extends the execution time of big load operations by changing frequency and voltage of variable voltage processors. Many researches have been going on to control the energy consumption of the processors and the data transmission among processors by controlling the bandwidth to reduce the energy consumption of the entire system. Since operations in multiprocessor systems have the data dependency between processors, however, the DVS techniques devised for single processors are not suitable to improve the energy efficiency of multiprocessor systems. We propose a new scheduling algorithm based on DVS for increasing energy efficiency of multiprocessor systems. The proposed DVS-based scheduling algorithm can improve the energy efficiency of entire systems because it controls frequency and voltages while considering the data dependency among processors.
Key concepts: Dynamic voltage scaling, Computer science, Multiprocessing, Energy consumption, Frequency scaling, MPSoC, Scheduling (production processes), Parallel computing