pd-gem5: Simulation Infrastructure for Parallel/Distributed Computer Systems for Network-Driven Optimization

Abstract

In developing and optimizing a parallel/distributed computer system, it is critical to study the impact of the com-plex interplay amongst processor, node, and network architectures on performance and power efficiency in de-tail. This necessitates a flexible, detailed, and open-source full-system simulation infrastructure, but our commu-nity lacks such an infrastructure. Responding to such a need, we present pd-gem5, a gem5-based infrastructure that can model and simulate a parallel/distributed computer system using multiple simulation hosts. We show that pd-gem5 running on 6 simulation hosts speeds up the simulation of a 24-node computer system up to 3.7? com-pared to simulating all 24 nodes in a single simulation host. As a use case of pd-gem5, we model a parallel/distributed computer system after enabling pd-gem5 with the Linux default ondemand governor. While running parallel/distributed workloads, we observe that a sudden in-crease in network activity often immediately leads to and/or is highly correlated with a high utilization of processors receiving the packets. However, the governor does not instantly increase the frequency of the processors. Such delayed responses often result in violations of service level agreements (SLAs), discouraging service providers from deploying an aggressive power management governor. To achieve both short response time and high energy-efficiency, we propose a network-driven power management governor and evaluate its efficacy compared with the default on-demand and performance governors.