Energy Minimization for Parallel Real-Time Systems with Malleable Jobs and Homogeneous Frequencies

TL;DR

This paper explores energy-efficient scheduling of malleable parallel jobs on multiprocessors, demonstrating that fixed frequencies can be optimal and proposing a polynomial-time algorithm that significantly reduces power consumption.

Contribution

It introduces a polynomial-time optimal processor and frequency selection algorithm for malleable jobs, showing dynamic voltage/frequency scaling is unnecessary for optimality.

Findings

Up to 60 watt power savings over non-parallel approaches

Fixed frequency scheduling can be optimal for energy minimization

Algorithm validated through hardware-based simulations

Abstract

In this work, we investigate the potential utility of parallelization for meeting real-time constraints and minimizing energy. We consider malleable Gang scheduling of implicit-deadline sporadic tasks upon multiprocessors. We first show the non-necessity of dynamic voltage/frequency regarding optimality of our scheduling problem. We adapt the canonical schedule for DVFS multiprocessor platforms and propose a polynomial-time optimal processor/frequency-selection algorithm. We evaluate the performance of our algorithm via simulations using parameters obtained from a hardware testbed implementation. Our algorithm has up to a 60 watt decrease in power consumption over the optimal non-parallel approach.