Engineers at the University of California, Santa Cruz made what researchers called a "small, but important step" forward for artificial intelligence. A circuit of roughly 100 artificial synapses was proved it had the ability to classify images.

The feat is something humans and animals do daily without giving much thought to it. However, with a rudimentary proof of concept now established, the researchers backed up the belief that artificial neural network could someday be built out to resemble, or best, the one quadrillion synaptic connections that make up the human brain.

"While the circuit was very small compared to practical networks, it is big enough to prove the concept of practicality," stated study author Merrikh-Bayat. "And, as more solutions to the technological challenges are proposed, the technology will be able to make it to the market sooner."

The neural network was able to classify three letters: "Z," "V" and "N" though each letter was distorted with noise.

The circuitry of the neural network is composed of "memristors," a fusion of memory and resistors. Memristors differ from conventional transistors in that they transfer of ions rather than electrons, allowing them to function more in line with the way human cells pass along electrical signals.

While traditional transistors lose power, even briefly, they also lose all of the information they were carrying. Memristor retain the profiles of the charges that pass through it, according to Dmitri Strukov, professor of electrical and computer engineering.

"For example, many different configurations of ionic profiles result in a continuum of memory states and hence analog memory functionality," said Strukov. "Ions are also much heavier than electrons and do not tunnel easily, which permits aggressive scaling of memristors without sacrificing analog properties." 

While the progress in neural networks has been promising, there's still a long way to go before super intelligent machines rise to over throw human civilization. Again, their artificial neural network was described as containing approximately 100 synapses. The human brain is composed of about one quadrillion synapses --- that's a 1 followed by 15 zeros.

For the near term, memristors could still push computing ahead by leaps and bounds.

"The exciting thing is that, unlike more exotic solutions, it is not difficult to imagine this technology integrated into common processing units and giving a serious boost to future computers," said Mirko Prezioso, lead researcher.

ⓒ 2024 TECHTIMES.com All rights reserved. Do not reproduce without permission.
Join the Discussion