NEUROMORPHIC COMPUTERS
Imagine how exciting it would be to work on a super computer which works more like a human brain.Steve ferber has made this impossible thing turn into reality .
A million core version of machine that will eventually be able to simulate upto billions of neurons per second. (We would take a short overview about the nerve cells which are known as neurons - these are located in the central nervous system which transmits impulses /information to every part of the body and vice versa )
Each of million cores can simulate roughly a millions synapses(space between two neurons)
So depending on the model 1000 neurons with 1000 connections or 100 neurons with 10000 connections.
To give an idea of scale, a mouse brain consists of around 100 million neurons and the human brain is 1,000 times bigger than that.
(One billion neurons is one per cent of the scale of the human brain.)
These computers are capable of completing more than 200 million million actions per second .
Now let's try to understand the working of this computer's in a simpler way .
Traditional computers that we use in daily life , thinks in a binary manner i.e everything is either in 1 and 0 or yes and no.Traditional computers work in a very rigid way whereas neurological computer's work in a flexible way. This computing involves the engineering and physics used in our own human brains. NEUROMORPHIC IS AN HARDWARE VERSION. In neurological computer's instead of using electrical signals to mean 1 or 0 , designers of this new chip's have made their computes neurons talk to each other the way biological neurons do ! Precise electric current is passed to the synapse (space between computing neurons) , depending upon the number and kind of ions receiving computing neurons gives you more computational options than a just basic yes and no.This ability to transfer information from "neurons to neurons",and all these processes working together simultaneously ,that means that NEUROMORPHIC chips could be more energy efficient than our normal computers.More amount of work is done within less energy .In these computers silicon material is used to get well controlled flow of ions (energy)between the neurons.
Adding more to your knowledge ....
A team in Korea is investigating other materials. They used tantalum oxide to give them precise control over the flow of ions and it's even more durable.
Another team in Colorado is using magnets to precisely cantrol the way computer neurons communicate.All these teams are working to get us to place where the neurons on the chip can " LEARN AS THEY COMPUTE".
Another Leap in the performance has been made by a team in Manchester who has taken a different approach which is called as spiNNaker which stands for spiking neural network agriculture .The hope is that the computer which works like a human brain can make us understand one of the most complicated thing "THE BRAIN ".
This is one of the biggest invention of today's life .
I would come with more blogs which will give you a short glance on a new inventions in the tech industry.
Visit this link for more information on spiNNaker .
0 Comments