Synaptic CMOS transistor
Neuromorphic systems inspired by the brain have gained much popularity in the computing world for its potential energy-efficient computation. To fully realisse a developed neuromorphic hardware application, large-scale integration of artificial synapses onto a single computing chip is required. In t...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
Nanyang Technological University
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/139089 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Neuromorphic systems inspired by the brain have gained much popularity in the computing world for its potential energy-efficient computation. To fully realisse a developed neuromorphic hardware application, large-scale integration of artificial synapses onto a single computing chip is required. In this work, electrical measurements were conducted on commercial level TiN/HfO2 gated n-channel MOSFET transistor (with an equivalent oxide thickness of 1.7nm) to demonstrate that it can exhibit some critical synaptic characteristics of a biological synapse such as excitatory and inhibitory postsynaptic current (EPSC and IPSC), short-term plasticity (STP) and long-term potentiation (LTP), metaplasticity, and spike timing dependent plasticity (STDP). The mechanism behind the output characteristic of MOSFET to exhibit synaptic-like response is the charge trapping and de-trapping at defects in the oxide and oxide/semiconductor interface. The readily available CMOS transistor can be a potential fundamental building block for an artificial neural network to drive towards a commercialised neuromorphic system. |
|---|