Brain like computing should significantly reduce energy consumption

It's an exciting new one Researchs area that is part of the megatrend Artificial intelligence (AI) is becoming increasingly important: the so-called Neuromorphic computingAlso Brain like computingg employs technology that mimics the human brain and nervous system. With this it could solve extensive associative learning problems. In addition, there is an opportunity to significantly reduce the energy consumption of the current silicon-based circuits. This works in the EU project Neuronn, which started in 2020 Fraunhofer EMFT.

The research team is working with six European partners on a new neuromorphic approach based on energy-efficient elements and architectures. In the proposed neurologically inspired computer architecture, information is encrypted in the phase of coupled oscillating elements that are interconnected to form a neural network.

Predictive maintenance with artificial intelligence for robots

Similar to the brain, the two key components in neuromorphic computing are called neurons and synapses. They simulate the distributed computing and storage units. Novel elements based on vanadium dioxide are used as neurons in the Brain Like Computing project. These are 250 times more efficient than state-of-the-art digital oscillators based on CMOS can be.

It's all about the synapses

At EMFT, the work package focuses specifically on the synapses. As part of the Brain-like Computing project Neuronn, the scientists are developing 2D nanoscale memristors based on innovative 2D nanomaterials. The tiny components are said to be 330 times more efficient than currently used technologies in terms of service life, switching speed and energy consumption.

The 36-month project brings together leading European academic and research institutions. They are Fraunhofer EMFT, IBM Research Zurich, CSIC / University of Seville, AI Mergence, FR and Silvaco UK. The French Center National De La Recherche Scientifique CNRS is responsible for the coordination. In addition, Neuronn has initiated an industrial advisory board consisting of members of the Intel Corporation and Prophesee.

Kick-off in early February

At the beginning of February 2020, the kick-off meeting of Neuronn took place in Montpellier, France at Lirmm, CNRS. The project is funded under the EU research program Horizon 2020 under grant number 871501.