A High Swing and Low Power Associative Memory Based on Emerging Technologies

Abstract

Over time, the need to process complex tasks has increased. This need leads to intelligent processors. This paper proposed a robust nonvolatile associative memory based on spintronic synapses utilizing magnetic tunnel junction (MTJ) and carbon nanotube field-effect transistors (CNTFET)-based neurons. The Proposed synapse produces more weights with more distance between them. The proposed design uses the MTJ device because of its fascinating nonvolatility feature. At the same time, CNTFET has overcome the conventional CMOS shortcomings like the short channel effect and low hole mobility. The proposed design is simulated in the presence of process variations. The proposed design in this paper produces 2.2 times more weights with the same number of MTJs compared to a similar design, and this advantage has a high impact on the capacity of the implemented associative memory.