Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Approximate Computing Survey, Part II: Application-Specific & Architectural Approximation Techniques and Applications

Vasileios Leon, Muhammad Abdullah Hanif, Giorgos Armeniakos, Xun Jiao, Muhammad Shafique, Kiamal Pekmestzi, Dimitrios Soudris

TL;DR

This survey addresses the growing need to reduce energy and latency in compute- and memory-intensive AI/DSP workloads by exploring Approximate Computing (AxC) across application-specific and architectural dimensions. It classifies and details software-, hardware-, and cross-layer techniques, offering a quantitative view of their trade-offs and highlighting representative gains (e.g., substantial energy savings with minimal accuracy loss) in diverse domains. The work emphasizes cross-layer co-design, end-to-end optimization, and the necessity of validated hardware models, benchmarks, and evaluation metrics to enable practical deployment. By surveying architecture-oriented accelerators, approximate memories, and data representations alongside application-driven methods, the paper charts a path for scalable, energy-efficient systems in edge and data-center contexts and outlines open challenges and future directions for AxC research.

Abstract

The challenging deployment of compute-intensive applications from domains such as Artificial Intelligence (AI) and Digital Signal Processing (DSP), forces the community of computing systems to explore new design approaches. Approximate Computing appears as an emerging solution, allowing to tune the quality of results in the design of a system in order to improve the energy efficiency and/or performance. This radical paradigm shift has attracted interest from both academia and industry, resulting in significant research on approximation techniques and methodologies at different design layers (from system down to integrated circuits). Motivated by the wide appeal of Approximate Computing over the last 10 years, we conduct a two-part survey to cover key aspects (e.g., terminology and applications) and review the state-of-the art approximation techniques from all layers of the traditional computing stack. Part II of the survey classifies and presents the technical details of application-specific and architectural approximation techniques, which both target the design of resource-efficient processors/accelerators and systems. Moreover, it reports a quantitative analysis of the techniques and a detailed analysis of the application spectrum of Approximate Computing, and finally, it discusses open challenges and future directions.

Approximate Computing Survey, Part II: Application-Specific & Architectural Approximation Techniques and Applications

TL;DR

This survey addresses the growing need to reduce energy and latency in compute- and memory-intensive AI/DSP workloads by exploring Approximate Computing (AxC) across application-specific and architectural dimensions. It classifies and details software-, hardware-, and cross-layer techniques, offering a quantitative view of their trade-offs and highlighting representative gains (e.g., substantial energy savings with minimal accuracy loss) in diverse domains. The work emphasizes cross-layer co-design, end-to-end optimization, and the necessity of validated hardware models, benchmarks, and evaluation metrics to enable practical deployment. By surveying architecture-oriented accelerators, approximate memories, and data representations alongside application-driven methods, the paper charts a path for scalable, energy-efficient systems in edge and data-center contexts and outlines open challenges and future directions for AxC research.

Abstract

The challenging deployment of compute-intensive applications from domains such as Artificial Intelligence (AI) and Digital Signal Processing (DSP), forces the community of computing systems to explore new design approaches. Approximate Computing appears as an emerging solution, allowing to tune the quality of results in the design of a system in order to improve the energy efficiency and/or performance. This radical paradigm shift has attracted interest from both academia and industry, resulting in significant research on approximation techniques and methodologies at different design layers (from system down to integrated circuits). Motivated by the wide appeal of Approximate Computing over the last 10 years, we conduct a two-part survey to cover key aspects (e.g., terminology and applications) and review the state-of-the art approximation techniques from all layers of the traditional computing stack. Part II of the survey classifies and presents the technical details of application-specific and architectural approximation techniques, which both target the design of resource-efficient processors/accelerators and systems. Moreover, it reports a quantitative analysis of the techniques and a detailed analysis of the application spectrum of Approximate Computing, and finally, it discusses open challenges and future directions.
Paper Structure (24 sections, 3 figures, 10 tables)

This paper contains 24 sections, 3 figures, 10 tables.

Table of Contents

  1. Introduction
  2. Approximate Computing Survey
  3. Application-Specific Approximation Techniques
  4. Software-Level Approximations
  5. Pruning
  6. Precision Scaling
  7. Input-Adaptive Approximation
  8. Input Scaling
  9. Hardware-Level Approximations
  10. Cross-Layer and End-to-End Approximations
  11. Architectural Approximation Techniques
  12. Approximate Processors
  13. Approximate Data Storage
  14. Approximate Non-Volatile Memories
  15. Approximate Volatile Memories
  16. ...and 9 more sections

Figures (3)

  • Figure 1: Number of publications that apply any type of approximation in 4 major design automation conferences.
  • Figure 2: Organization of the proposed two-part survey on Approximate Computing.
  • Figure 3: Classification of application-specific approximation techniques for DL and AI applications in three main classes: Software-Level, Hardware-Level, and Cross-Layer & End-to-End.