TU Dresden
A joint project with Uni Leipzig DE EN

2D-Material-Based Synapses for Neuromorphic Computing

Mentored by Thomas Mikolajick, Martin Bogdan
at TU Dresden, Institute of Semiconductors and Microsystems, Chair of Nanoelectronics

Neuromorphic computation is considered as one of the most promising approaches to overcome the von Neumann bottleneck as well as the scaling limitations of CMOS devices and enable very energy efficient computing systems of he future. First neuromorphic systems like the Loihi chip from INTEL are still constructed using state of the art CMOS technology and, therefore, can only partially exploit the full potential of neuromorphic computing systems.

In the last years, attempts are made to construct new devices that can realize the functions of neurons and synapses with a lower device count, smaller footprint and much higher energy efficiency. Examples are synapses made from phase change devices, resistive switching devices or ferroelectric devices as well as neurons based on ferroelectric field effect transistors or threshold switching devices. In future neuromorphic systems, 3-dimensional architectures will be required to achieve the required density for very complex tasks. Therefore, the device functionality should be independent of the substrate, e.g. silicon. From the other end, in the last years 2D materials have demonstrated to have excellent electrical properties and can be explored to realize various functionalities like transistors, resistive switches or sensors. In this project, synaptic functions using either resistive switching or ferroelectric switching in 2D materials will be realized and tuned to achieve the best fir towards the required characteristics from the system point of view.

Work Environment

You will be working with Prof. Dr.-Ing Thomas Mikolajick and Dr. Andre Heinzig and will have access to the TU Dresden cleanroom Lab the Joint 2D device Lab of TU Dresden and HZDR. You will be supported by Prof. Dr. Martin Bogdan to define the required device characteristics and receive feedback on the results from a system point of view especially based on the Modified Stochastic Synaptic Model (MSSM). Several research stays at the Leipzig University will be included in the project.

The chair of Nanoelectronics at TU Dresden runs a fully equipped state of the art research cleanroom for the fabrication of test structures and test devices. You will have access to the cleanroom, and will be responsible for the key processes used during the work. Just recently the cleanroom equipment was updated in the framework of the project FoRLab DCST. Moreover, the chair runs a new lab setup to process and integrated 2D materials together with the Helmholtz Center Dresden Rossendorf (HZDR) that will be used by the student pursuing this project. Finally, Labs for electrical and structural characterization are available either at the chair or via collaboration partners.

Prerequisites

To conduct this research, you should hold a very good university degree (MSc or an equivalent) in Electrical Engineering, material Science Physics or related disciplines (such as applied natural sciences, electronic and senor materials, nanoelectronics, microsystem technology or other specialized programs in the field). Basic knowledge on semiconductor technology and semiconductor devices is required and first experience in working in a cleanroom as well as characterizing or simulating semiconductor devices are welcome but not absolutely required.