Power aware page allocation

Abstract

One of the major challenges of post-PC computing is the need to reduce energy consumption, thereby extending the lifetime of the batteries that power these mobile devies. Memory is a particularly important target for efforts to improve energy efficiency. Memory technology is becoming available that offers power management features such as the ability to put individual chips in any one of several different power modes. In this paper we explore the interaction of page placement with static and dynamic hardware policies to exploit these emerging hardware features. In particular, we consider page allocation policies that can be employed by an informed operating system to complement the hardware power management strategies. We perform experiments using two complementary simulation environments: a trace-driven simulator with workload traces that are representative of mobile computing and an execution-driven simulator with a detailed processor/memory model and a more memory-intensive set of benchmarks (SPEC2000). Our results make a compelling case for a cooperative hardware/software approach for exploiting power-aware memory, with down to as little as 45% of the Energy Delay for the best static policy and 1% to 20% of the Energy Delay for a traditional full-power memory.