Reducing data movement energy via online data clustering and encoding

Modern computer systems expend significant amounts of energy on transmitting data over long and highly capacitive interconnects. A promising way of reducing the data movement energy is to design the interconnect such that the transmission of 0s is considerably cheaper than that of 1s. Given such an interconnect with asymmetric transmission costs, data movement energy can be reduced by encoding the transmitted data such that the number of 1s in each transmitted codeword is minimized. This paper presents a new data encoding technique based on online data clustering that exploits this opportunity. The transmitted data blocks are dynamically clustered based on the similarities between their binary representations. Each data block is expressed as the bitwise XOR between one of multiple cluster centers and a residual with a small number of 1s. The data movement energy is minimized by sending the residual along with an identifier that specifies which cluster center to use in decoding the transmitted data. At runtime, the proposed approach continually updates the cluster centers based on the observed data to adapt to phase changes. The proposed technique is compared to three previously proposed energy-efficient data encoding techniques on a set of 14 applications. The results indicate respective energy savings of 5%, 9%, and 12% in DDR4, LPDDR3, and last level cache subsystems as compared to the best existing baseline encoding technique.

I. Introduction

Data movement is a major contributor to the total system energy consumption in deeply scaled CMOS ICs [1]. Studies show that for scientific and mobile applications, 40% [2] and 35% [3] of the total system energy is consumed by data movement, respectively. The energy cost of data movement is substantially higher than that of computation. For example, when performing a double precision addition on a graphics processing unit (GPU) implemented at the 22nm technology node, fetching the two operands from memory consumes greater energy than moving the operands from the edge of the chip to its center, which in turn consumes another higher energy than the addition [4]. This orders of magnitude energy gap between data movement and computation is expected to widen in the future with technology scaling [5]. Thus, reducing data movement energy is critical to future computer systems.