Multi-Objective Application-Driven Approximate Design Method

Salvatore Barone,Mario Barbareschi,Marcello Traiola,Alberto Bosio

doi:10.1109/access.2021.3087858

Abstract

Approximate Computing (AxC) paradigm aims at designing computing systems that can satisfy the rising performance demands and improve the energy efficiency. AxC exploits the gap between the level of accuracy required by the users, and the actual precision provided by the computing system, for achieving diverse optimizations. Various AxC techniques have been proposed so far in the literature at different abstraction levels from hardware to software. These techniques have been successfully utilized and combined to realize approximate implementations of applications in various domains (e.g. data analytic, scientific computing, multimedia and signal processing, and machine learning). Unfortunately, state-of-the-art approximation methodologies focus on a single abstraction level, such as combining elementary components (e.g., arithmetic operations) which are firstly approximated using component-level metrics and then combined to provide a good trade-off between efficiency and accuracy at the application level. This hinders the possibility for designers to explore different approximation opportunities, optimized for different applications and implementation targets. Therefore, we designed and implemented E- \mathbb IDEA, an automatic framework that provides an application-driven approximation approach to find the best approximate versions of a given application targeting different implementations (i.e., hardware and software). E- \mathbb IDEA compounds 1) a source-to-source manipulation tool and 2) an evolutionary search engine to automatically realize approximate application variants and perform a Design-Space Exploration (DSE). The latter results in a set of non-dominate approximate solutions in terms of trade-off between accuracy and efficiency. Experimental results validate the effectiveness and the flexibility of the approach in generating optimized approximate implementations of different applications, by using different approximation techniques and different accuracy/error metrics and for different implementation targets.

Highlights

Approximate Computing (AxC) term has been introduced to define a computing paradigm applied at different abstraction levels, spanning from hardware to software components [1]
2) RESULTS In Figure 3, we report the percentage of skipped loops (y-axis) and the corresponding SSIM index (x-axis) for the non-dominated solutions found with E-IDEA
B: RESULTS In Figure 7, we report the percentage of skipped loops (y-axis) and the corresponding SSIM index (x-axis) for the non-dominated solutions found with E-IDEA

Summary

Introduction

Approximate Computing (AxC) term has been introduced to define a computing paradigm applied at different abstraction levels, spanning from hardware to software components [1]. For several real-world scenarios, the effectiveness of imprecise computation has been demonstrated in the literature, for both software and hardware components implementing. The inherent resiliency property tightly depends on the application domain. It can be observed for algorithms dealing with noisy real-world input data (e.g., image processing, sensor data processing, speech recognition, etc.), or with outputs that have to be interpreted by humans, such as digital signal processing of images or audio; data analytic, web search and wireless communications exhibit an equivalent property [4]–[6]. Other involved applications are those that iteratively process large amounts of information, sample data, stop the convergence procedure early, or apply heuristics [7]

Objectives

Results

Conclusion