Semiconductor MemorySemiconductor Memory for Neuromorphic Computing


Current computer is a fast computing machine, not a thinking device like our brain capable of intelligent recognition. Some brain functions such as speech recognition have often been emulated by computers, but the way they are implemented on computers is inherently different from the way the brain realizes these functions. The brain is a vast array of neurons and synapses. Each neuron receives stimuli from adjacent neurons connected through synapses; once a neuron is adequately stimulated, it in turn stimulates (or fires) other neurons that it connects to. The way a set of neurons is connected through synapses and how they stimulate other neurons determines a particular function.

  Our research goal is to discover the best way of realizing neurons and synapses, and how to extract the connections and firing rules for each particular target function. This can be done collaboration including memristors (synapses), devices and circuits (neurons), algorithms (connections and firing rules), and systems (target functions). Especially, we focused on memristor based neuromorphic computing system, emulating the potentiation and depression process based on identical pulses called Spike-timing dependent plasticity (STDP). Our memristor-based synapse show gradual resistance transition between Low Resistance State (LRS) to High Resistance State (HRS) which is essential for multi-level cell (MLC) behavior needed for complex neuromorphic computing.

[Related References]

"Novel Electronics for Flexible and Neuromorphic Computing", Adv. Funct. Mater.  28, 1801690, 2018

"Unconventional Memristive Devices for Advanced Intelligent Systems", Adv. Mater. Technol., 4, 1900080, 2019  

"Reliable Memristive-Switching Memory Devices Enabled by Densely-Packed Silver Nanocone Arrays as Electric-Field Concentrators" ACS Nano, 10, 9478, 2016. 

"Self-Structured Conductive Filament Nanoheater for Phase Change Memory" ACS Nano, 9, 6587, 2015.

"Reliable Control of Filament Formation in Resistive Memories by Self-assembled Nano-insulators Derived from a Block Copolymer"  ACS Nano, 8, 9492, 2014,  

"Flexible Crossbar Resistive Memory Arrays via Inorganic-based Laser Lift-off "  Adv. Mater., 26, 7480, 2014

"Self-assembled Incorporation of Modulated Block Copolymer Nanostructures in Phase-change Memory for Switching Power Reduction", ACS Nano, 7(3), 2651, 2013.  

"Flexible Memristive Memory Array on Plastic Substrates", Nano Letters, 11(12) , 5438, 2011.

이용약관 ㅣ개인정보처리방침

Department of Materials Science and Engineering, KAIST ㅣ Fax: 82-42-350-3310 ㅣ TEL: 82-42-350-3343 ㅣAddress : 291 DaeHak-ro, Yuseong-gu, Daejeon, Korea, 34141 (대전 유성구 대학로 291)

Copyright  © 2021. KAIST. All rights reserved.

Department of Materials Science and Engineering, KAIST

Fax: 82-42-350-3310 ㅣ TEL: 82-42-350-3343 ㅣ

Address : 291 DaeHak-ro, Yuseong-gu, Daejeon, Korea, 34141 (대전 유성구 대학로 291)

Copyright  © 2021.KAIST. All rights reserved.