THE HUMAN BRAIN : Energy efficient, compact and the ultimate computer
DAWN OF A NEW AGE : limitations of existing hardware
Since the dawn of the computer age researchers and scientists alike have been trying to improve the computer and get it to function more efficiently and more purposefully. The computer has evolved over the years and with artificial intelligence just across the horizon things are looking good. However, there are certain limitations that are incarcerating the further development of the computer system for instance the ever increasing processor clock speeds are simply unsustainable and the heavier RAMs especially in Windows driven systems limit battery time. In a world where everybody and everything is on the go computers are expected to give increasingly good battery times but the heavy spec hardware that these machines carry makes it nearly impossible to go beyond a certain limit with regards to good battery times. And what if we downgrade the hardware? With the average user demanding more computational power than ever before downgrading the hardware may not be the best idea…OR IS IT??
IBM SyNAPSE ®™
“We now have the seeds of a new architecture that can allow us to mine the boundary between the physical and the digital world in an ever more efficient way.” – Dharmendra Modha IBM Research
IBM has designed prototype chips that explicitly mimic the human brain in terms of the way the hardware circuit is set up. Now instead of having those regular “buses” to carry electric currents to various parts of the computer, these chips form a dual core processor named True North which utilizes 256 “digital neurons” running at slow speeds of 10MHz that are constantly blasting information to each other. The two cores contain roughly about 350,000 digital synapses of which 250,000 are programmable. Like in the brain, the synapse establishes connections between digital neurons, and the more often a signal is sent to a synapse, the stronger the synapse gets which is very similar to how the “buses” work in your regular laptops/desktop computers.
From these chips IBM hopes to build a fully functional computer system that will utilize 10 billion neurons and 100 trillion synapses with the power consumption and size that rivals the human brain. KEY PHRASE: LOW ON POWER CONSUMPTION and so maybe a brain mimicking laptop would be the perfect portable device which will cater to our “always on the go” needs.
*SyNAPSE is a backronym standing for Systems of Neuromorphic Adaptive Plastic Scalable Electronics. The name alludes to synapses, the junctions between biological neurons. The program is being undertaken by HRL Laboratories (HRL), Hewlett-Packard, and IBM Research.
THE MEMRISTOR AND NEUROMORPHIC CHIPS
The memristor has been touted as the artificial electronic equivalent of the synapse present in the human brain and consists of fine nanolayers of circuitry. Synapses in the brain work on impulses and then carry out the best course of action and the thing about these synapses is that they learn this course of action and carry out the same course of action when a similar impulse is presented to them. A memristor works on a similar blueprint and instead of impulses of course, a computer system uses electric currents. So, the memristor would record the amount of current that flows through it and the amount of current that flows out of it – this ability can be molded into introducing the learning capability into systems.
NEUROMORPHIC CHIPS : A sneak peek into the future
Neuroinformatics researchers from the University of Zurich and ETH Zurich together with colleagues from the EU and US set out to go one step further than IBM’s SyNAPSE ®™ and introduce the ability to process data like a human mind whereas the former’s main focus is to produce an energy efficient system. These researchers exhibited how cognitive capabilities of the human brain can be integrated into complex electronic systems by the use of these chips which they’re calling “neuromorphic chips”. The main component of these chips is of course the memristor which sort of sounds like a distant cousin of the word memory so no surprises as to what its specialty is.
Now these neuromorphic chips possess the ability to process data and produce the best possible output with the added feature that they will try and produce similar responses for similar data entries – this is very similar to how we have preferences with regards to stuff like our favorite clothes, food etc. So this roughly translates into a low power system that has the ability to LEARN how cool is that??
I’ll give you a rough demonstration : Imagine one day these chips power your smartphones so on a fine sunday morning you have a bowl of apples in front of you and you’re not sure if they’re fresh or not so you flip out your phone, powered by a neuromorphic chip, and you scan these apples using the camera app and boom ! this is what it’ll do