Executive Summary : | The electronic industry has made significant progress in reducing the cost, size, and power requirements of chips through scaling as per Moore's Law. From 1956 floating-point operations to the 2016 supercomputer (RIKEN), the power cost of floating-point operations has been reduced by almost 400 billion. This has allowed the traditional Von-Neumann computing paradigm to survive, but it has led to increased energy consumption and wasted time. Several approaches have been proposed for beyond Von-Neumann architecture, such as neuromorphic computing or spintronic device-based processors. Spintronic devices, such as Magnetic Tunnel Junction (MTJ) devices, can work as both a logic and memory cell, offering integrated memory architecture beyond the Von-Neumann bottleneck. These devices are commonly used in MRAM, HDD read heads, and logic applications. The spin-based computing system offers features such as non-volatility, high speed operation, zero static power, good scalability, high density, and lesser device count per logic gate. The challenge for beyond Von-Neumann architecture lies in developing materials and devices that can permit this paradigm. Two-dimensional (2D) materials have been found to efficiently manipulate, transport, and generate spin signals, making them a promising platform for building spintronic devices beyond the CMOS/Von-Neumann architecture. |