A group of physicists from the University of Lodz is investigating whether it is possible to construct a computer from two-dimensional materials, i.e., materials with a thickness of one or several atoms. This is to enable construction of memory systems similar to the human brain. The project worth over PLN 2 million is financed through the National Science Centre.
Research focuses on a phenomenon called resistive switching. We talk about this phenomenon when the electrical resistance changes as electricity flows through a given material.
Resistive switching is similar to the way the synapses that connect neurons in the human brain work. Information repeatedly flowing between neurons can strengthen the synapses connecting them and facilitate the process of association, just as the current flowing through resistive switching systems can increase their conductivity – explains dr Maciej Rogala from the Faculty of Physics and Applied Informatics at the University of Lodz.
In recent years, it has been shown that resistive switching can occur faster and more efficiently than other physical processes that are currently used to record information and that it can revolutionize the data carrier market. At the same time, the use of resistive switching to build neuromorphic systems would mean a complete change in the approach towards computers construction, in which, like in the human brain, information would be processed directly in the memory instead of being constantly transferred between the memory and processor.
It seems extremely advantageous to use two-dimensional materials for the purpose of resistive switching, i.e., materials the thicknesses of which is at the level of individual atoms. The best known two-dimensional material is graphene. However, this group, as shown by the discoveries of the last decade, is much wider.
Thanks to such materials, it is possible to build memory and computing systems that are extremely thin (almost transparent) or extremely densely packed and at the same time dissipate heat very efficiently. The first attempts at building resistive switching systems based on two-dimensional materials have yielded very promising results. However, to take full advantage of their capabilities, we need to understand the basic physical and chemical mechanisms that drive resistive switching processes.
Such processes in two-dimensional materials take place on the atomic scale and on such a scale they will be analyzed in this project. In the future this will allow for the creation of optimal artificial synapses and efficient hardware neural networks, which may be a revolution in the information processing methods.
Source: Faculty of Physics and Applied Informatics, UL (Wydział Fizyki i Informatyki Stosowanej, UŁ)