Memristor Modeling with COMSOL Multiphysics®

Memristors represent a key advancement in nonvolatile memory technology and neuromorphic computing. These devices are characterized by their ability to retain resistance states based on past voltage and current history, mimicking synaptic behavior in biological systems. This unique feature makes memristors particularly well suited for applications in artificial neural networks, offering more efficient data storage and processing compared to traditional transistor-based systems.

Current applications extend across several domains, including memory storage solutions, edge computing, advanced sensing systems, and adaptive learning algorithms. Their unique characteristics not only pave the way for innovative digital architectures but also revolutionize how data is processed and stored, marking a significant leap toward nonconventional computing systems.

In this webinar, we will welcome Prof. Valeria Bragaglia, a physicist who specializes in condensed matter, focusing on developing materials and devices for next-generation computing in the Neuromorphic Devices & Systems lab at IBM Research Europe - Zurich. Prof. Bragaglia will share insight into the fundamentals of memristors and how her team has used the COMSOL Multiphysics^® software to successfully develop redox-based resistive switching random access memory (ReRAM) devices.

Donato Francesco Falcone, a PhD student in the same team at IBM, soon to graduate from ETH Zurich, will then take the audience through the key electrothermal challenges in designing memristive devices. Mr. Falcone will discuss the building blocks essential for solving these challenges with COMSOL^® and the development of design rules for optimizing next-generation ReRAM devices.

In the second part of the webinar, there will be a live demo introducing the basic steps of a memristor simulation with COMSOL^®.