Artificial Neural Network


Current efforts to overcome the problems of slow speed of operation and high consumption of power for computers are focused on the downsizing of device, but this is rapidly becoming impossible to continue due to constraints in the working size of material and ability of lithographic tool to resolve the detail in a device. To improve the performance of device, a completely different way of computing is needed. Here, we explore new methods to overcome the problems in speed and power of a device by controlling temporary glass state of a material. Applying repeated electrical pulses in a well organised manner not only enables fast, low power and efficient (or parallel) computing, but also allows neural-network-like programming, which could potentially benefit from a dynamic control of the temporary glass states.


• Artificial neural network using temporary glass states

• Computer simulation of artificial neural networks


1. T. H. Lee, D. Loke, K. J. Huang, W. J. Wang, S. R. Elliott. Tailoring Transient-Amorphous States: Towards Fast and Power-Efficient Phase-Change Memory and Neuromorphic Computing. Adv. Mater. 26, 7493-7498 (2014).

2. J. M. Skelton, D. Loke, T. H. Lee, S. R. Elliott. Ab Initio Molecular-Dynamics Simulation of Neuromorphic Computing in Phase-Change Memory Materials. ACS Appl. Mater. Interfaces 7, 14223-14230 (2015).