Microprocessor’s Power Consumption Optimization Using Combined Pareto optimal Solutions and Particle Swarm Optimization Algorithms

Abstract

Power consumption becomes an important design constraint for high-performance systems and microprocessor's designs, It is considered as a key constraint when making design decisions. High power/temperatures and eminent thermal variations of processors pose serious challenges for system performance and reliability. It shortens CPU lifetime and raise packaging and cooling costs. As a result, power and thermal management issues have emerged as key factors in the design of high-performance processors, which limit the performance of microprocessors, portable and computer systems. In this article, a multi-objective particle swarm optimization (MOPSO) approach is presented for generating Pareto-Optimal Solutions (POS) for microprocessor’s power consumption and temperature variations problems. This method is proposed and developed by integrating Pareto dominance principles into Particle Swarm Optimization (PSO) algorithm. The obtained results confirm the theoretical basics for efficient high-performance microprocessors, and the power consumption and temperatures were minimized successfully.

Keywords

• power consumption
• temperature variations
• pareto front (pf)
• particle swarm optimization (pso)