Exploring the potential of neuromorphic computing

While computers have grown exponentially more powerful over the last few decades, they still fall short of the most powerful computer in the world: the human brain. Neuromorphic computing devices aim to replicate the functionality of the brain, with the goal of replicating its performance as well.

Mondal et al. discussed recent advances in developing neuromorphic devices, potential applications of these emerging technologies, and directions for future research.

“Unlike traditional computers, which store data in one place and process it in another, neuromorphic systems combine these functions in a single place, much like neurons in the brain do,” said author Giridhar Kulkarni. “This approach allows these systems to use less energy, process enormous amount of information in better and faster ways, and even learn from past experiences.”

The authors highlighted the development of emerging memory and processing technologies that can mimic the synaptic plasticity found in neurons, giving them the ability to adapt their responses based on previous stimuli. They also provided an overview of neuromorphic designs implemented by chip manufacturers.

Looking to the future, the researchers called for more consistent results from neuromorphic systems and for improving their miniaturization and energy efficiency. They are optimistic that new materials and designs along with continued support from semiconductor manufacturers will allow neuromorphic technologies to reach their full potential.

“Imagine robots that can adapt to new situations or medical devices that can better understand brain signals — these are the kinds of advancements we’re looking forward to,” said Kulkarni. “The potential for smarter, more intuitive technology is what makes this field so exciting.”

Source: “Recent trends in neuromorphic systems for non-von Neumann in materia computing and cognitive functionalities,” by Indrajit Mondal, Rohit Attri, Tejaswini S. Rao, Bhupesh Yadav, and Giridhar U. Kulkarni, Applied Physics Reviews (2025). The article can be accessed at https://doi.org/10.1063/5.0220628 .