Abstract
Improving the performance and power efficiency of a single processor has been fraught with various challenges stemming from the end of the classical technology scaling. Thus, the importance of efficiently running applications on a parallel/distributed computer system has continued to increase. In developing and optimizing such a parallel/distributed computer system, it is critical to study the impact of the complex interplay amongst processor, node, and network architectures on performance and power efficiency in detail. This necessitates a flexible, detailed and open-source full-system simulation infrastructure. However, our community lacks such an infrastructure. In this paper, we present pd-gem5, a gem5-based infrastructure that can model and simulate a parallel/ distributed computer system using multiple simulation hosts. Our experiment shows that pd-gem5 running on six simulation hosts speeds up the simulation of a 24-node computer system up to 3.2× compared with running on a single simulation host.
Keywords
Parallel/distributed systems, Simulation infrastructure, Power efficiency, Performance optimization.