Dynamic Scheduling-Based Optimization of Multi-Core Architectures for High-Performance Computing

Abstract

This study aims to solve the problem of task scheduling and load balancing in multi-core architectures and proposes an optimization method based on dynamic scheduling strategy to improve the efficiency and resource utilization of high-performance computing systems. In traditional multi-core systems, as the number of cores increases, the optimization of task scheduling and resource allocation becomes more and more complicated. Especially when dealing with computationally intensive and memory-intensive tasks, how to maximize the computing power of multi-core processors becomes a key issue. This paper experimentally analyzes the performance of different scheduling strategies (static scheduling, dynamic scheduling, and priority scheduling) under different core number configurations. The results show that compared with other methods, the dynamic scheduling strategy can optimize execution efficiency and resource utilization while improving system throughput. In addition, the experiment also explores the trend of performance improvement after the increase in the number of cores and the possible performance saturation phenomenon, revealing the impact of scheduling strategies on the overall performance of the system in large-scale multi-core systems. The research results provide a theoretical basis for the design and optimization of future multi-core computing architectures and point out the direction for the development of intelligent scheduling algorithms. Future work will further study adaptive scheduling strategies in heterogeneous computing environments, combined with artificial intelligence and deep learning methods to achieve more efficient resource management and optimization solutions.