The team has built a small chip with artificial synapses using silicon germanium. In simulations, the researchers found that the chip and its synapses could be used to recognize samples of handwriting, with 95 percent accuracy.
Most neuromorphic chip designs attempt to emulate the synaptic connection between neurons using two conductive layers separated by a switching medium. When a voltage is applied, ions move in the switching medium to create conductive filaments, similarly to how the weight of a synapse changes.
In this architecture is it difficult to control the flow of ions in existing designs as there are many possible paths. “Once you apply some voltage to represent some data with your artificial neuron, you have to erase and be able to write it again in the exact same way,” said Jeehwan Kim, asst professor in the departments of Mechanical Engineering and Materials Science and Engineering, and a principal investigator in MIT’s Research Laboratory of Electronics and Microsystems Technology Laboratories. “But in an amorphous solid, when you write again, the ions go in different directions because there are lots of defects,” he said. “This stream is changing, and it’s hard to control. That’s the biggest problem — nonuniformity of the artificial synapse.”
Instead of using amorphous materials as an artificial synapse, Kim and his colleagues used single-crystal silicon and created a precise, one-dimensional line defect, or dislocation, through the silicon, through which ions could predictably flow.
The team then grew a lattice of silicon germanium on top of the silicon wafer. Silicon germanium’s lattice is slightly larger than that of silicon, and Kim found that together, the two perfectly mismatched materials can form a funnel-like dislocation, creating a single path through which ions can flow.
The neuromorphic chip consists of artificial synapses made from silicon germanium (SiGe), each around 25nm in size. The team applied voltage to each synapse and found that all synapses exhibited more or less the same current, or flow of ions, with about a 4 percent variation between synapses — a much more uniform performance compared with synapses made from amorphous material.
They also tested a single synapse over multiple trials, applying the same voltage over 700 cycles, and found the synapse exhibited the same current, with just 1 percent variation from cycle to cycle. “This is the most uniform device we could achieve, which is the key to demonstrating artificial neural networks,” said Kim.
As a final test, Kim’s team explored how its device would perform if it were to carry out actual learning tasks such as recognizing samples of handwriting. The team ran a computer simulation of an artificial neural network consisting of three sheets of neural layers connected via two layers of artificial synapses with the properties of the actual neuromorphic chip. They fed into their simulation tens of thousands of samples from a handwritten recognition dataset commonly used by neuromorphic designers, and found that their neural network hardware recognized handwritten samples 95 percent of the time, compared to the 97 percent accuracy of existing software algorithms.
The team is in the process of building a working neuromorphic chip that can carry out handwriting-recognition tasks. Beyond handwriting, the artificial synapse design will enable much smaller, portable neural network devices that can perform complex computations that currently are only possible with large supercomputers.
“Ultimately we want a chip as big as a fingernail to replace one big supercomputer,” said Kim. “This opens a stepping stone to produce real artificial hardware.”