A Simulation Framework for Memristive Deep Learning Systems

This repository includes the Resistive Random Access Memory (RRAM) Compiler which is designed in the context of the research project of Dimitris Antoniadis (PG Taught Student) at Imperial College London

Python tools for generating and testing SPICE netlists/waveforms involving crossbar memory arrays in various configurations

A well-posed RRAM SPICE model implemented in Verilog-A, based on Stanford/ASU filamentary model, using code developed at UC Berkeley

Series of ReRAM characterization modules compatible with Keithley Semiconductor Characterization Systemsinstruments

Code and repository for RRAM RADAR programming method: https://doi.org/10.1109/TED.2021.3097975

Long Short-Term Memory Implementation Exploiting Passive RRAM Crossbar Array

Memristor model: Various implementations of the simplified memristor model "JART-TUD VCM"

NI RRAM programming in Python

Scripts to model functional experimental or other phenomena, such as neuronal/device spiking, or tip-sample interactions.

RRAM resistive switching behavior evaluation and prediction, based on fabrication conditions. Applied Machine learning\Deep learning models to predict SET voltage distribution in Honey-based RRAM devices. Simulating the basic operations of RRAM crossbars in image classification tasks, investigating the robustness of in situ vs. ex situ training.