PARE: A Power-Aware Data Prefetching Engine

Publication Files

Publication Medium:

International Symposium on Low Power Electronics and Design

pages

pp. 339 - 344

Year of Publication:

2005

Abstract

Aggressive hardware prefetching often significantly increases energy consumption in the memory system. Experiments show that a major fraction of prefetching related energy degradation is due to the hardware history table related energy costs. In this paper, we present PARE, a PowerAware pRefetching Engine that uses a newly designed indexed hardware history table. Compared to the conventional single table design, the new prefetching table consumes 7-11X less power per access. With the help of compilerbased location-set analysis, we show that the proposed PARE design improves energy consumption by as much as 40% in the data memory systems in 70-nm BTPM processor designs.

Research Category