This is Unikorn, an attempt to determine a new quantum dot based transistor prototype.
This project uses Ab-Initio and DFT simulation methods, coupled with a message passing neural network machine learning model and an automatic optimization algorithm to automatically optimize and search for a new transistor based on maximizing both logical throughput (ternary level logic) and speed (carrier transport). It is hoped that utilizing technologies such as MMMMM (Metropolis Monte-Carlo), and speeding up DFT calculations with Iridium (MPNN Accelrator) will allow for a more efficient and accurate search of the molecule parameter space.
Some of these are private.
Neural Network
Monte-Carlo Simulation Platform
The Library for Simulation
Full Project
Some of these such as the blog may be of interest to you.
Homepage of the Owner
Owner's Blog
Email