Scientists at the University of Zurich and ETH Zurich, together with colleagues in Germany and the United States, created electronic systems comparable to a human brain both in size, speed and energy consumption, the university said in a statement late Monday.
Just like the brain, their so-called neuromorphic chips are capable of processing and reacting to information in real-time, it said.
“The challenge is to build something as close as possible to an actual brain,” Giacomo Indiveri, a University of Zurich professor of Neuroinformatics and one of the researchers on the project, told AFP.
Electronic systems in the past have been designed to react to their environments, as with blinds that automatically close when sunlight hits them.
But, said Indiveri, the new project takes things further.
Using neuromorphic chips as artificial neurons, the researchers built networks that can perform tasks requiring short-term memory and decision-making and analytical abilities, Indiveri said.
The technology could over time become a useful tool, allowing robots to “navigate autonomously in an environment and survive without someone with a remote control,” he said, adding that the chips might also help make smart phones even smarter.
The chips could also one day pave the way for computers that can function despite faulty parts, in the same way the human brain continues to churn unabated even though it loses around a million neurons each day.
The findings are published in the US journal Proceedings of the National Academy of Sciences (PNAS).