(2DoNeuron) 2D photonic memristive devices for neuromorphic applications

What is 2DoNeuron about?

2DoNeuron aims to develop electrical and optically controllable artificial neural networks based on photonic, low-power consumption and high integration density 2D-based memristors. This will allow to directly train optical memristors to classify standard handwritten digit images as a first proof-of principle case study.

Our vision is that the assembly of such optical adaptive switches, based on rich 2D materials properties, into novel architecture systems will provide a revolutionary paradigm for heavily distributed, ultrafast data processing with vast impact in neuromorphic computing.

By successfully accomplishing the fabrication of an optical neural network we will give a significant step to pave the way for real-time processing of information encoded by light patterns.

What critical challenges is 2DoNeuron addressing?

The pursuit for new device architectures and novel materials is essential for the next generation computing technology. Accordingly, memristors have attracted increasing attention due to their unique properties and great potential in neuromorphic computing.

Moreover, if light can be used as an extra control parameter in these memristors, it is attractive for remote controlled memory and artificial synapses. Therefore, one needs optically active materials meeting the demands of the active memristor layer, which makes novel two-dimensional (2D) materials promising for multifunctional optoelectronic computing and memory.

How will 2DoNeuron optimize the production of neuromorphic applications?

Our strategy will start with the fabrication of optimized 2D materials-based memristive devices, profiting from the extended know-how and experience of the partner groups on resistive switching behaviors and 2D-materials fabrication. We will use a combination of appropriate materials, growth conditions and lithography processes.

The memristive devices will display neuromorphic properties such as the extended Hebb rule and STDP. The operation parameters will be studied both under electrical and optical stimulation. This is a new field, with 2D-based memristors still in their infancy. To take full advantage of 2D materials, the fabrication should be significantly optimized for integration with optical systems, and it is very important to understand the underlying switching mechanisms.

How is 2DoNeuron contributing to nanomaterial research?

Devices with such rich properties, would serve as foundation for building biorealistic neuromorphic computing systems, opening new applications in photonics, including optical switches for integrated (on-chip) photonic circuits, optical communication and new types of light sensors that can be used in cellular neural network cameras. These results will be the foundation for the longer-term objectives to allow the science-to-technology breakthrough of these optical artificial neural networks to reach the market. This immediately promotes Portugal’s international competitiveness and innovation capacity.