Energy-Aware Thread Assignment and Resource Allocation in Multi-Server Systems

In modern computing environments, distributed systems have become the key architecture for handling large-scale computational tasks. Effective thread assignment and resource allocation are crucial to achieve energy-efficient system. This paper investigates the scientific problem of joint optimization of thread assignment and resource allocation in distributed systems for energy saving. By analyzing the processes of thread assignment and resource allocation, we propose an efficient optimization framework to enhance system performance. We introduce the Multi-server Energy Consumption Optimization (MECO) algorithm, which aims to find the optimal combination by calculating the latency increments caused by different threadserver combinations. Simulation experiments demonstrate that our algorithm reduces the energy consumption by up to 65.99 %, compared to several other algorithms.