Computational architecture that mimics the brain could revolutionize processors

Researchers at the University of Limerick, Ireland, in partnership with the National University of Singapore and Texas A&M University, in the USA, have developed a new computing architecture inspired by the human brain and formed by a molecule composed of just 77 atoms.

The device works as memory and processor simultaneously, providing a new electronic element that can be reconfigured in real time to perform predefined logic functions. The system is also able to “learn” new information by mimicking brain synapses.

“Similar to the flexibility and adaptability of connections in the human brain, our memory device can be quickly reconfigured for different computational tasks simply by changing the applied electrical voltages. In addition, just like nerve cells, this system can retain information for future retrieval and processing,” explains Singapore National University engineering professor Sreetosh Goswami, co-author of the study.

Want to catch up on the best tech news of the day? Access and subscribe to our new youtube channel, Canaltech News. Everyday a summary of the main news from the tech world for you!


The new computational architecture, inspired by the internal activities of the brain, was created by optimizing the electrical properties of soft crystals produced from a single molecule that belongs to the chemical family of phenyl-azo-pyridines. This new electronic circuit is more robust and is not limited to storing information from the binary system that varies between 0 and 1, as occurs in transistors.

New computational architecture consists of just one molecule (Image: Reproduction/University of Limerick)

This molecule uses its natural asymmetry in metal-organic bonds to switch between different states, enabling lightning-fast decision processing in real time. “They function like electron sponges, which can provide up to six electron transfers, resulting in five different molecular states,” adds Goswami.

Unlike conventional transistors, made with metal oxides turned on and off with fixed voltages, the new molecular components can switch between these states using several sequential voltages. This means that a processing unit will not need to fetch data for each calculation operation, saving time and energy.


Since a single molecular memory circuit can perform the same amount of computational functions as thousands of transistors put together, tiny identical processors could be networked together and run in parallel, allowing for troubleshooting even if the individual components don’t work correctly all the time. .

Molecular memory working scheme (Image: Reproduction/University of Limerick)

This redundancy and reconfigurability of the system show that the new device, which mimics the human brain, can do everything in one place, without the need to read and move information all the time, allowing for the development of more complex and efficient electronic circuits.

“The new circuit elements can provide smaller, faster and more energy-efficient computers, exactly what is needed for high-end computing, internet of things and artificial intelligence applications,” concludes the University of Limerick physics professor Damien Thompson, lead author of the study.

Source: University of Limerick

Did you like this article?

Subscribe your email on Canaltech to receive daily updates with the latest news from the world of technology.

Related Articles

Back to top button