Byung+Chul+Jang
-
KAIST Develops Analog Memristive Synapses for Neuromorphic Chips
(Professor Sung-Yool Choi from the School of Electrical Engineering)
A KAIST research team developed a technology that makes a transition of the operation mode of flexible memristors to synaptic analog switching by reducing the size of the formed filament. Through this technology, memristors can extend their role to memristive synapses for neuromorphic chips, which will lead to developing soft neuromorphic intelligent systems.
Brain-inspired neuromorphic chips have been gaining a great deal of attention for reducing the power consumption and integrating data processing, compared to conventional semiconductor chips. Similarly, memristors are known to be the most suitable candidate for making a crossbar array which is the most efficient architecture for realizing hardware-based artificial neural network (ANN) inside a neuromorphic chip.
A hardware-based ANN consists of a neuron circuit and synapse elements, the connecting pieces. In the neuromorphic system, the synaptic weight, which represents the connection strength between neurons, should be stored and updated as the type of analog data at each synapse.
However, most memristors have digital characteristics suitable for nonvolatile memory. These characteristics put a limitation on the analog operation of the memristors, which makes it difficult to apply them to synaptic devices.
Professor Sung-Yool Choi from the School of Electrical Engineering and his team fabricated a flexible polymer memristor on a plastic substrate, and found that changing the size of the conductive metal filaments formed inside the device on the scale of metal atoms can make a transition of the memristor behavior from digital to analog.
Using this phenomenon, the team developed flexible memristor-based electronic synapses, which can continuously and linearly update synaptic weight, and operate under mechanical deformations such as bending.
The team confirmed that the ANN based on these memristor synapses can effectively classify person’s facial images even when they were damaged. This research demonstrated the possibility of a neuromorphic chip that can efficiently recognize faces, numbers, and objects.
Professor Choi said, “We found the principles underlying the transition from digital to analog operation of the memristors. I believe that this research paves the way for applying various memristors to either digital memory or electronic synapses, and will accelerate the development of a high-performing neuromorphic chip.”
In a joint research project with Professor Sung Gap Im (KAIST) and Professor V. P. Dravid (Northwestern University), this study was led by Dr. Byung Chul Jang (Samsung Electronics), Dr. Sungkyu Kim (Northwestern University) and Dr. Sang Yoon Yang (KAIST), and was published online in Nano Letters (10.1021/acs.nanolett.8b04023) on January 4, 2019.
Figure 1. a) Schematic illustration of a flexible pV3D3 memristor-based electronic synapse array. b) Cross-sectional TEM image of the flexible pV3D3 memristor
2019.02.28 View 7849 -
Ultra-Low Power Flexible Memory Using 2D Materials
(Professor Choi and Ph.D. candidate Jang)
KAIST research team led by Professor Sung-Yool Choi at School of Electrical Engineering and Professor Sung Gap Im at the Department of Chemical and Biomolecular Engineering developed high-density, ultra-low power, non-volatile, flexible memory technology using 2D materials. The team used ultrathin molybdenum disulfide (MoS2) with atomic-scale thickness as the channel material and high-performance polymeric insulator film as the tunneling dielectric material. This research was published on the cover of Advanced Functional Materials on November 17. KAIST graduate Myung Hun Woo, a researcher at Samsung Electronics and Ph.D. candidate Byung Chul Jang are first authors.
The surge of new technologies such as Internet of Things (IoT), Artificial Intelligence (AI), and cloud server led to the paradigm shift from processor-centric computing to memory-centric computing in the industry, as well as the increase in demand of wearable devices. This led to an increased need for high-density, ultra-low power, non-volatile flexible memory. In particular, ultrathin MoS2 as semiconductor material has been recently regarded as post-silicon material. This is due to its ultrathin thickness of atomic-scale which suppresses short channel effect observed in conventional silicon material, leading to advantages in high- density and low-power consumption. Further, this thickness allows the material to be flexible, and thus the material is applicable to wearable devices.
However, due to the dangling-bond free surface of MoS2 semiconductor material, it is difficult to deposit the thin insulator film to be uniform and stable over a large area via the conventional atomic layer deposition process. Further, the currently used solution process makes it difficult to deposit uniformly low dielectric constant (k) polymeric insulator film with sub-10 nm thickness on a large area, thus indicating that the memory device utilizing the conventional solution-processed polymer insulator film cannot be operated at low-operating voltage and is not compatible with photolithography.
The research team tried to overcome the hurdles and develop high-density, ultra-low power, non-volatile flexible memory by employing a low-temperature, solvent-free, and all-dry vapor phase technique named initiated chemical vapor deposition (iCVD) process. Using iCVD process, tunneling polymeric insulator film with 10 nm thickness was deposited uniformly on MoS2 semiconductor material without being restricted by the dangling bond-free surface of MoS2. The team observed that the newly developed MoS2-based non-volatile memory can be operated at low-voltage (around 10V), in contrast to the conventional MoS2-based non-volatile memory that requires over 20V.
Professor Choi said, “As the basis for the Fourth Industrial revolution technologies including AI and IoT, semiconductor device technology needs to have characteristics of low-power and flexibility, in clear contrast to conventional memory devices.” He continued, “This new technology is significant in developing source technology in terms of materials, processes, and devices to contribute to achieve these characteristics.”
This research was supported by the Global Frontier Center for Advanced Soft Electronics and the Creative Materials Discovery Program by funded the National Research Foundation of Korea of Ministry of Science and ICT.
( Figure 1. Cover of Advanced Functional Materials)
(Figure 2. Concept map for the developed non-volatile memory material and high-resolution transmission electron microscopy image for material cross-section )
2018.01.02 View 8014