Certain investigations on performance enhancement of memristor based in memory computing architecture

Abstract

Due to the rapid development in the semiconductor industry, there newlineare various challenges in digital VLSI system design, particularly in data newlineintensive applications like medical imaging, autonomous car navigation, space newlineexploration, and underwater scenarios. This research highlights the challenges newlinefaced in modern VLSI system design, including high off-chip memory access newlinelatency, limited memory bandwidth, and significant data movement overhead, newlinewith delayed processing time and energy utilization. Energy efficiency, area, newlinedelays, and sneak path current are primary concerns, especially in portable newlineelectronic devices. A multiplication operation in an arithmetic unit is a newlinesignificant source of power dissipation in digital VLSI circuits. To address newlinethese challenges, the research focuses on developing efficient low-power newlinearithmetic circuits by approximate computing technique to reduce area and newlinedelay, specifically addressing power dissipation during idle states. newlineThe design of approximate computing architecture includes the newlinedesign of an approximate adder, its integration into multiplier circuits, and the newlineoptimization of memristor-based memory arrays. This adder introduces newlineminimal errors to outputs, significantly reducing power consumption while newlinepreserving accuracy. By integrating approximate computing techniques and newlinememristor-based memory solutions, the research seeks to improve the newlineperformance of digital VLSI systems across various applications including newlineimage processing, data storage and computation. Memristor-based In-memory newlinecomputing architecture is required to optimize memory elements and enhance newlineimage quality. newline