Minimizing the Power Consumption of a Chip Multiprocessor System under an Average Throughput Constraint

Mohammad Ghasemazar,  Ehsan Pakbaznia,  Massoud Pedram
University of Southern California


Abstract

Designers are moving toward Chip Multiprocessors (CMPs) due to the increasing demand for higher processing power. In a multi-core system, power and performance may be dynamically traded off by utilizing power management (PM). This paper addresses the problem of minimizing the total power consumption of a chip multiprocessor system while maintaining a target average throughput. The proposed solution relies on a hierarchical framework, which employs core consolidation, coarse-grain dynamic voltage and frequency scaling (DVFS), and task assignment at the CMP level and fine-grain DVFS based on closed-loop feedback control at the individual core level. Experimental results show an average of 23% power saving compared to a baseline technique, and demonstrate the high efficacy of the proposed hierarchical PM framework.