New electronics mimic brain's low-power computing abilities

April 20 (UPI) -- Engineers have developed a new electronic device that mimics the brain's synapses. The miniature technological tools, called memristors, send electric signals across protein nanowires with unprecedented efficiency.

Scientists described the new device, a so-called neuromorphic memristor, or "memory transistor," in a new paper published Monday in the journal Nature Communications.

The most efficient computers require at least a volt of power to send and process information. Meanwhile, the brain needs just 80 millivolts to send electric signals, or action potentials, between neurons.

For engineers trying to approximate the computing power of the brain, one of the most daunting obstacles has been the disparity in energy efficiency. The latest research suggests that obstacle is not insurmountable.

In the lab, the new memristor, built using protein nanowires derived from the bacterium Geobacter, achieved neurological voltages.

"This is the first time that a device can function at the same voltage level as the brain," study co-author Jun Yao, electrical and computer engineering researcher at the University of Massachusetts Amherst, said in a news release. "People probably didn't even dare to hope that we could create a device that is as power-efficient as the biological counterparts in a brain, but now we have realistic evidence of ultra-low power computing capabilities. It's a concept breakthrough and we think it's going to cause a lot of exploration in electronics that work in the biological voltage regime."

The Geobacter-derived nanowires offer significant conductive advantages, but they're also much more eco-friendly. The production of traditional silicon nanowires is energy-intensive and requires the use of toxic chemicals. And because protein nanowires are more stable in water and bodily fluids, they can be more easily integrated into biomedical technologies.

To test the device's efficiency, scientists combined the protein nanowires with a metal thread to form a simple electric switch. When scientists sent a pulsing charge pattern through the switch, they found the nanowires work by augmenting the thread's metal ion reactivity and electron transfer qualities. The changes in the metal thread recalled the kind of machine learning deployed by the brain.

"You can modulate the conductivity, or the plasticity, of the nanowire-memristor synapse so it can emulate biological components for brain-inspired computing," Yao said. "Compared to a conventional computer, this device has a learning capability that is not software-based."

The results of early experiments with the memristor-powered electric switch weren't all that impressive, but over time, they saw significant improvement's the device's efficiency.

"I remember the day we saw this great performance," said first author Tianda Fu, a doctoral candidate in electrical and computer engineering. "We watched the computer as current voltage sweep was being measured. It kept doing down and down and we said to each other, 'Wow, it's working.' It was very surprising and very encouraging."

The research team plans to continue conduction experiments using protein nanowires and to incorporate memristors into medical devices like a heart rate monitor.

"This offers hope in the feasibility that one day this device can talk to actual neurons in biological systems," Yao said.