Project description
A 'brain' is born of novel nanomaterials with emergent neuronal properties
The most simplistic computational model of a neuron is an 'on-off' switch, with a '0' representing a resting state and a '1' representing an axon firing an action potential. While this lends itself well to conventional digital electronics and silicon-based transistors, it does not represent the incredible natural 'state' space of a real neuron. When it comes to realising the potential of a brain-like processing system, novel materials are needed. The EU-funded MELON project has created an expert consortium of academic institutions and an SME to explore novel materials with history-dependent conductivity to emulate neuronal connectivity. Together with materials capable of multivalued logic and interconnects, the team plans to deliver the building blocks of tomorrow's emergent computing circuits.
Objective
To make a machine think like a human we should overcome the tyranny of the deterministic binary logic, inherent to the contemporary electronic circuits. While the realization of this emergent approach has long been suggested as a multi-valued and neuromorphic architecture of the logic units, the problem is that we haven’t discovered a material system that could implement it. Right now, silicon-based transistors can operate as “on” and “off”, so the new materials would have to find to consistently maintain more states and emulate the plasticity and self-organization of neuronal connections. It’s against this background we develop the Consortium within the RISE action “MELON”, involving academic members from EU Member States, France, Netherlands, and Spain, from the partner country, Argentina, and the SME from the associated country, Ukraine, with the objective to develop the innovative materials platform for the realization of the emergent computing circuits. We target three focus areas:
(i) to explore novel memristive oxide-based systems on silicon, with history-dependent conductivity, for emulating the neuronal connections in the brain,
(ii) to use the nano-scale multiferroic materials, hosting the multistable topological states to realize the multi-valued logic, and
(iii) to explore conducting 2D oxide interfaces as both novels 4-points memristive systems and interconnect elements for the computing circuits.
The Consortium combines the complementary expertise spanned from fundamental to applied chemistry and physics and from material science to (industrial or modern) nanotechnologies development with the solid interdisciplinary and intersectoral potential for skills transfer, staff exchange, and young researchers' training.
Fields of science
Keywords
Programme(s)
Coordinator
80025 Amiens
France