Tag Archives: neuromorphic chips

Artificial Synapses Made from Nanowires

Scientists from Jülich together with colleagues from Aachen and Turin have produced a memristive element made from nanowires that functions in much the same way as a biological nerve cell. The component is able to both save and process information, as well as receive numerous signals in parallel. The resistive switching cell made from oxide crystal nanowires is thus proving to be the ideal candidate for use in building bioinspired “neuromorphic” processors, able to take over the diverse functions of biological synapses and neurons.

Image captured by an electron microscope of a single nanowire memristor (highlighted in colour to distinguish it from other nanowires in the background image). Blue: silver electrode, orange: nanowire, yellow: platinum electrode. Blue bubbles are dispersed over the nanowire. They are made up of silver ions and form a bridge between the electrodes which increases the resistance.

Computers have learned a lot in recent years. Thanks to rapid progress in artificial intelligence they are now able to drive cars, translate texts, defeat world champions at chess, and much more besides. In doing so, one of the greatest challenges lies in the attempt to artificially reproduce the signal processing in the human brain. In neural networks, data are stored and processed to a high degree in parallel. Traditional computers on the other hand rapidly work through tasks in succession and clearly distinguish between the storing and processing of information. As a rule, neural networks can only be simulated in a very cumbersome and inefficient way using conventional hardware.

Systems with neuromorphic chips that imitate the way the human brain works offer significant advantages. Experts in the field describe this type of bioinspired computer as being able to work in a decentralised way, having at its disposal a multitude of processors, which, like neurons in the brain, are connected to each other by networks. If a processor breaks down, another can take over its function. What is more, just like in the brain, where practice leads to improved signal transfer, a bioinspired processor should have the capacity to learn.

With today’s semiconductor technology, these functions are to some extent already achievable. These systems are however suitable for particular applications and require a lot of space and energy,” says Dr. Ilia Valov from Forschungszentrum Jülich. “Our nanowire devices made from zinc oxide crystals can inherently process and even store information, as well as being extremely small and energy efficient,” explains the researcher from Jülich’s Peter Grünberg Institute.

Source: http://www.fz-juelich.de/

Brain function partly replicated by nanomaterials

The brain requires surprisingly little energy to adapt to the environment to learn, make ambiguous recognitions, have high recognition ability and intelligence, and perform complex information processing.

The two key features of neural circuits are “learning ability of synapses” and “nerve impulses or spikes.” As brain science progresses, brain structure has been gradually clarified, but it is too complicated to completely emulate. Scientists have tried to replicate brain function by using simplified neuromorphic circuits and devices that emulate a part of the brain’s mechanisms.

Spontaneous spikes being similar to nerve impulses of neurons was generated from a POM/CNT complexed network

In developing neuromorphic chips to artificially replicate the circuits that mimic brain structure and function, the functions of generation and transmission of spontaneous spikes that mimic nerve impulses (spikes) have not yet been fully utilized.

A joint group of researchers from Kyushu Institute of Technology and Osaka University studied current rectification control in junctions of various molecules and particles absorbed on single-walled carbon nanotube (SWNT), using conductive atomic force microscopy (C -AFM), and discovered that a negative differential resistance was produced in polyoxometalate (POM) molecules absorbed on SWNT. This suggests that an unstable dynamic non-equilibrium state occurs in molecular junctions.

In addition, the researchers created extremely dense, random SWNT/POM network molecular neuromorphic devices, generating spontaneous spikes similar to nerve impulses of neurons.

POM consists of metal atoms and oxygen atoms to form a 3-dimensional framework. Unlike ordinary organic molecules, POM can store charges in a single molecule. In this study, it was thought that negative differential resistance and spike generation from the network were caused by nonequilibrium charge dynamics in molecular junctions in the network.

Thus, the joint research group led by Megumi Akai-Kasaya conducted simulation calculations of the random molecular network model complexed with POM molecules, which are able to store electric charges, replicating spikes generated from the random molecular network.  They also demonstrated that this molecular model would very likely become a component of reservoir computing devices. Reservoir computing is anticipated as next-generation artificial intelligence (AI). Their research results were published in Nature Communications.

The significance of our study is that a portion of brain function was replicated by nano-molecular materials. We demonstrated the possibility that the random molecular network itself can become neuromorphic AI,” says lead author Hirofumi Tanaka.

Source: http://resou.osaka-u.ac.jp/