In a research breakthrough, scientists have developed a transistor that is capable of performing energy-efficient associative learning at room temperature. This new ‘brain-like’ transistor is described as being capable of mimicking a level of human intelligence. This is a type of memory resistor, or “memristor,” a device that can perform combined processing and memory function.
The transistor can simultaneously processes and stores information like the human brain, operating at room temperature, unlike previous devices which required cryogenic condition. This makes the new device more practical than earlier iterations. The device is constructed from atomically thin materials: bilayer graphene and hexagonal boron nitride.
This development is in keeping with advances in artificial intelligence (AI) that have motivated researchers to develop computers that operate more like the human brain.
To test the transistor, the researchers trained it to recognize similar — but not identical — patterns. The basis of this is because if AI is meant to mimic human thought, one of the lowest-level tasks to achieve is to classify data, which is sorting data into bins.
The new device is synaptic transistor capable of higher-level thinking and it comes from Northwestern University, Boston College, and the Massachusetts Institute of Technology. To demonstrate this, the researchers showed that the transistor goes beyond simple machine-learning tasks to categorize data. Instead, the transistor is capable of performing associative learning whilst operating at fast speeds.
Not only does the device consume very little energy it also retains stored information even when power is removed. The low power consumption has come about through advances in moiré patterns. This is a type of geometrical design that arises when two patterns are layered on top of one another.
Here, when two-dimensional materials are stacked, new properties emerge that do not exist in one layer alone. And when those layers are twisted to form a moiré pattern, more advanced tunability of electronic properties becomes possible.
Outlining why the new device is different to what has been developed previously, one of the lead researchers, Mark C. Hersam, states: “The brain has a fundamentally different architecture than a digital computer. In a digital computer, data move back and forth between a microprocessor and memory, which consumes a lot of energy and creates a bottleneck when attempting to perform multiple tasks at the same time.”
He adds: “On the other hand, in the brain, memory and information processing are co-located and fully integrated, resulting in orders of magnitude higher energy efficiency. Our synaptic transistor similarly achieves concurrent memory and information processing functionality to more faithfully mimic the brain.”
The study has been published in the journal Nature. The research is titled “Moiré synaptic transistor with room-temperature neuromorphic functionality.”
