FORAY-GEN: Automatic Generation of Affine Functions for Memory Optimizations

TL;DR

This paper introduces FORAY-GEN, an automated profiling tool that extracts a model of memory behavior from programs, enabling more effective energy optimization techniques in embedded systems.

Contribution

It presents a novel automated profiling method for extracting the FORAY model, improving the applicability of memory optimization approaches.

Findings

Doubles the number of memory references analyzable by static methods.

Enables aggressive memory behavior analysis in programs with loops and array accesses.

Facilitates energy-efficient memory subsystem design in embedded applications.

Abstract

In today's embedded applications a significant portion of energy is spent in the memory subsystem. Several approaches have been proposed to minimize this energy, including the use of scratch pad memories, with many based on static analysis of a program. However, often it is not possible to perform static analysis and optimization of a program's memory access behavior unless the program is specifically written for this purpose. In this paper we introduce the FORAY model of a program that permits aggressive analysis of the application's memory behavior that further enables such optimizations since it consists of 'for' loops and array accesses which are easily analyzable. We present FORAY-GEN: an automated profile-based approach for extraction of the FORAY model from the original program. We also demonstrate how FORAY-GEN enhances applicability of other memory subsystem optimization approaches, resulting in an average of two times increase in the number of memory references that can be analyzed by existing static approaches.