Energy-Efficient and Power-Aware Parallel Execution

Abstract

Author: Abhay Kumar Dhawan

Abstract: The rapid growth of parallel computing platforms, ranging from multi-core processors to large-scale data centers and heterogeneous accelerators, has significantly improved computational performance. However, this performance growth has been accompanied by a sharp increase in energy consumption and power density, making energy efficiency a primary design concern. Energyefficient and power-aware parallel execution focuses on reducing energy usage while maintaining acceptable performance levels. This paper presents a comprehensive review of techniques and strategies used to achieve energy efficiency in parallel execution environments. It discusses architectural-level innovations, operating system support, compiler optimizations, runtime management, and application-level approaches. The trade-offs between performance, energy consumption, and power constraints are analyzed. Case studies and comparative tables are provided to highlight the effectiveness of different approaches. The paper aims to serve as a useful reference for researchers and practitioners working on sustainable and energy-aware parallel computing systems.

Keywords: Energy efficiency, Power-aware computing, Parallel execution, Multi-core processors, Dynamic voltage and frequency scaling, Heterogeneous systems