Technological Hardware

Engineers Are Using Artificial Brain Synapses …

MIT engineers designed a “brain on a chip” that resembles a human brain, smaller than a piece of confetti made up of tens of thousands of artificial brain synapses – silicon-based components that mimic information-transmitting synapses. After the researches, they took the example of metallurgical principles to produce each memrix of silver and copper alloys with silicon.

A chip is an electronic circuit placed on a semiconductor metal sheet made of silicon consisting of an integrated circuit, chip, microchip, integrated circuit. The chip has been reproduced with clear and clean versions to un-stored images and memristor designs made with unalloyed elements.

Artificial synapse networks are available as software. An attempt is made to create real neural network hardware for portable artificial intelligence systems. “Imagine connecting a neuromorphic device to a camera in your car and recognizing lights and objects and making decisions right away without having to connect to the internet,” said Jeehwan Kim, associate professor of mechanical engineering at MIT. We hope to use energy-efficient memristors to do these tasks in real-time on the spot. ”

Traveling Ions
Memristors or memory transistors are an important element in neuromorphic computing.

In a neuromorphic device, a memristor acts as a transistor in a circuit. Studies conducted will resemble a brain synapse, that is, the connection between two neurons. The synapse receives signals from one neuron in the form of ions and sends a corresponding signal to the next neuron.

The transistor has one of two values ​​by switching between 0 and 1. The signal it receives by transition has a certain power in the form of an electric current.

Memristor works along a gradient like a synapse in the brain. The directional derivative of the gradient scalar field has a vector field oriented towards the place where the increment is highest.

It depends on the signal it generates and the strength of the signal it receives. This allows a single memristor1 to have many values. Therefore, the two transistors have a much wider operating range.

Memories containing more than normal transistors and capacitors provide the visual classification of a complex equation and object. The value associated with a certain current strength with the brain synapse

Memristor Designs
Memristor designs say they work quite well in situations where the voltage stimulates a large conduction channel or heavy ion flow from one electrode to the next. However, these designs are less reliable when memories need to generate finer signals through thinner transmission channels.

The thinner the conduction channel and the lighter the flow of ions from one electrode to another, the more difficult it was for individual ions to stay together. Instead, they tend to move away from the group as dispersed in the setting.

As a result, it is difficult for the receiving electrode to reliably capture the same number of ions and therefore transmit the same signal when excited with a certain low current range.

Metallurgy Borrowing
The group of engineers found a way around this limitation by borrowing a technique from metallurgy, studying the science of melting metals into alloys, and their compound properties.

“Traditionally, metallurgists have been trying to add different atoms to an aggregate matrix to strengthen materials, and we thought we wouldn’t change atomic interactions in our memristor and add some alloying elements to control the movement of ions in our environment,” explained Jeehwan Kim.

Engineers often use silver as the material for the memristor’s positive electrode. Kim’s team looked through the literature to find an element they could combine with silver to effectively hold the silver ions together, allowing them to flow rapidly to other electrodes.

The team identified copper as the ideal alloying element, as it can bond with both silver and silicon.

 

source:http://news.mit.edu

2 thoughts on “Engineers Are Using Artificial Brain Synapses …

Leave a Reply

Your email address will not be published. Required fields are marked *