IISc team achieves leap in computing: 16,500 conductance states within a molecular film lead to brain-like processor

However advanced today's digital computers are, they use a switching device called a transistor/MOSFET working to do Boolean computing and storage. The massive achievement in present-day computing is due to the ability to pack these transistors in billions inside silicon chips measuring a few square inches. Such hundreds of these chips, called GPUs, enable a ChatGPT-like GenAI to generate knowledge-like information by storing terabytes of data in various types of memory devices, including semiconductor-type and magnetic-type.
Though they all together generate content like a subject expert, the technology is circuitous and roundabout with a lot of inefficiencies compared to how a human brain works. The simple reason is that whatever the fed analog data is, it is converted and stored in digital '1' and '0' format. Unlike the human brain, the memory and processing elements stay at a distance from each other, where it takes a lot of energy and time for each of them to work together.

Researchers who are working to replicate the human brain on a synthetic man-made system find these digital systems not a sustainable way. Instead, they are looking for devices that can store as well as process data. In their hunt for such technologies, they are evaluating every material's ability to store multiple states, so that a single device, instead of storing '1' and '0', can store an...