Abstract

Random-access memory (RAM), however volatile, is used in modern computing systems for high-speed data transport. Memristors are employed as storage instead of solid-state devices due to their non-volatility and lower energy needs. Memristor-based digital design has gained popularity for low-power, space-efficient VLSI design. The most important computational units for building high-speed processors is the arithmetic logic unit (ALU). Being aware that these processors’ space and time requirements are significant due to the ALU’s constraint, memristors are employed to get around these restrictions. Memristors are used in the design and simulation of the building components that make up an ALU. In addition to having a smaller footprint than others and reduced latency, it is a strong contender for ALU design. This work compares CMOS logic with an ALU’s performance in terms of average power and integral energy for all of the memristor-based ALU’s computational units. In this work, performance analysis of a memristor-based ALU has been carried out for high-speed processors, taking fuzzy metrics into account to characterise the behaviour uncertainty of the memristor. To offer a complete picture of the ALU’s performance under various circumstances, fuzzy logic is utilised to evaluate the average power and integral energy of the ALU. The reconfigurable FPGAs’ high-likelihood characteristics can be used to model memristors.