Executive Summary : | Artificial neural networks are computational models inspired by the function and structure of neural networks of the biological brain. They have found wide-ranging applications in diverse fields such as image and speech recognition, language processing, and autonomous vehicles. However, as the complexity of the neural network increases, so does the computational power required to train and execute it. This has led researchers to explore alternative hardware architectures that can overcome the limitations of conventional computing paradigms, such as the von Neumann architecture. One promising approach is neuromorphic computing, which seeks to emulate the way the brain processes information by leveraging the principles of parallelism, sparsity, and low power consumption. One key component of neuromorphic computing is the crossbar array, which is a matrix of memristive or memtransitor devices that can perform analog multiplication and addition operations, similar to the way synapses in the brain function. However, current implementations of crossbar arrays suffer from several limitations, including high power consumption, limited scalability, and compatibility issues with existing complementary metal oxide semiconductor (CMOs) technology. To overcome these challenges, the proposed project seeks to develop a new type of crossbar array based on two-dimensional (2D) materials, which have shown great promise in enabling low-power, high-performance electronics. The project builds on the previous research by our group and others in the field of 2D materials, which have been shown to exhibit unique electronic properties such as high electron mobility, tunable bandgap, and strong mechanical flexibility. By integrating 2D materials such as Mos2 and Ws2 into the crossbar array architecture, the researchers aim to achieve improved device performance, scalability, and compatibility with CMOs technology. The project will involve a range of research activities, including materials synthesis and characterization, device fabrication and testing, circuit design and optimization, and system-level integration and validation. The ultimate goal is to demonstrate the feasibility of the proposed crossbar array architecture for implementing artificial neural networks and neuromorphic computing applications, and to pave the way for the development of next-generation hardware platforms that can revolutionize the field of computing. This project is a research initiative aimed at developing a new type of hardware architecture for implementing artificial neural networks (ANNs) and neuromorphic computing applications. |