RoBA multiplier: A rounding-based approximate multiplier for high-speed yet energy-efficient digital signal processing

View at publisher Summarise with VIRAT
Authors: Chinthala Aruna, T. S. Arulananth, Manoj Kumar
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This paper introduces a Rounding-Based Approximate (RoBA) Multiplier that offers a practical solution to the increasing demand for energy-efficient and high-speed operations in Digital Signal Processing (DSP) applications. Unlike traditional multipliers that depend on exact calculations using full-width input values, the RoBA architecture takes a much simpler approach. It rounds the inputs to the nearest power of two, which cuts down the number of partial products and turns the multiplication into basic bitshifting and addition operations. This method not only helps reduce the size of the hardware but also allows the system to run faster and use less power. Based on what I studied, the RoBA Multiplier was built using 45nm CMOS technology and gave very promising results. It showed up to 45% energy savings and around 30% better speed when compared to exact multipliers. Even though the design uses approximation, the error remained low and acceptable for real-time DSP applications. The RoBA Multiplier a great choice for areas like multimedia processing, embedded systems, and edge computing where a small drop in accuracy is fine if it leads to better speed and energy efficiency.

Fields of Study & Subjects

Low-power high-performance VLSI design Numerical Methods and Algorithms Analog and Mixed-Signal Circuit Design

Publication details

Journal / source
CRC Press eBooks
Pages
540–547
Publication date
19 March 2026
Language
EN
Document type
Book chapter