The current chips used in computers drain power to perform the simplest operations, like adding or subtracting as they require the use of many transistors. They also can’t process information in parallel. This is where they differ from human brains, not to mention, they can’t recognize a visual pattern or understand human language.
Alternative chip use design that deviate from the traditional norms tackle all the above problems by using neuron-like structures or resorting to massive parallelism, but they still need human minds for shaping their software.
A research team led by professors Wilfried van der Wiel and Hajo Broersma from the University of Twente in the Netherlands have devised a new type of electronic chip that is massively parallel, and can manipulate data in arbitrary ways – even though it doesn’t need to be explicitly designed to perform any task, it takes after the human brain.
The chip itself decides the best path to follow and judges the outcome. This proof-of-principle device consists of a network of up to 100 densely interconnected gold nanoparticles, 20 nanometers in size, each acting as a tiny transistor. This product of artificial evolution is dubbed as the natural computer as the particles and transistors that make it up are are connected in a specific pattern, the shared network is exactly like our brain’s’ neuron connections.
The nanoparticle cluster contains two input signals, one output signal, and six control voltages that are connected to the end and determine the output of the network. The entire circuit is not handmade, but instead the researchers function by knowing the exact combination of control voltages that process all the inputs into the correct outputs.
“Although we understand the generic physics principles underlying the cluster’s behavior, we do not know the actual current paths in the network on a nanoscale level,” Van der Weil said.
This method will prove cheaper and more sustainable in the long run as if the chip were to suffer damage, then it could be recovered by adjusting the voltage values for the signals that control how the nanoparticle cluster manipulates the data. A genetic algorithm (GA) in computer science, is a clever way to optimize a set of parameters for an arbitrary condition. Selection of the optimal values for the six control voltages is done by this method which takes less than an hour.
The researchers were able to indirectly “program” the chip to act as any one of the Boolean logic gates by optimizing the voltages for different functionalities.
“Natural computers have, in general, the promise to be more energy efficient,” says Van der Weil. “The promise for lower energy consumption is based on the fact that natural computers take more advantage of the computational power of matter than conventional computers, and that many computational processes occur in parallel as opposed to sequentially as in conventional computers.”