Artificial synapse – one more step towards artificial intelligence
One of the biggest challenges for the research and development of artificial intelligence is to understand the human brain and figure out how to imitate it.
One such understanding has just been mimicked in an electronic circuit by He Tian and his colleagues from universities in Southern California and Florida, USA.
They developed an artificial synapse capable of simulate a fundamental function of our nervous system – the release of inhibitory and excitatory signals from the same “presynaptic” terminal.
The human nervous system consists of more than 100 trillion synapses, structures that allow neurons to emit electrical and chemical signals to each other. In mammals, these synapses can initiate and inhibit biological messages. Many synapses only transmit one type of signal, while others can transmit both types simultaneously, or they can switch between the two.
Most of the artificial synapses manufactured to date, however, – most based on the memristors – are only capable of firing one type of signal.
Reconfigurable artificial synapse
Now, He Tian has created a reconfigurable artificial synapse, capable of sending excitatory and inhibitory signals.
The synaptic component reconfigures itself based on the electrical voltages applied to its input terminal. A junction made of black phosphorus and tin selenide allows switching between excitatory and inhibitory signals.
The component is flexible and versatile, which is highly desirable for the fabrication of artificial neural networks.
With the presynaptic component, the expectation is that the design of artificial synapses and neuromorphic circuits will be simpler and may incorporate more functions, facilitating the construction of artificial intelligence in hardware.
Sources: Emulating Bilingual Synaptic Response Using a Junction-Based Artificial Synaptic Device
He Tian, Xi Cao, Yujun Xie, Xiaodong Yan, Andrew Kostelec, Don DiMarzio, Cheng Chang, Li-Dong Zhao, Wei Wu, Jesse Tice, Judy J. Cha, Jing Guo, Han Wang