scholarly journals Two-dimensional ferroelectric channel transistors integrating ultra-fast memory and neural computing

2020 ◽  
Author(s):  
Shuiyuan Wang ◽  
Lan Liu ◽  
Lurong Gan ◽  
Huawei chen ◽  
Xiang Hou ◽  
...  
Keyword(s):  
Energy Efficiency ◽  
Neural Plasticity ◽  
Iris Recognition ◽  
Computing System ◽  
Internal Electric Field ◽  
Fast Switching ◽  
Analog Computing ◽  
Non Volatile Memory ◽  
Non Destructive ◽  
With Memory

Abstract With the advent of big data era, applications are more data-centric, and energy efficiency issues caused by frequent data interactions due to the physical separation of memory and computing will become more severe. Emerging technologies have been proposed to perform analog computing with memory to address the dilemma. Ferroelectric memory has become a promising technology due to field-driven fast switching and non-destructive readout, but endurance and miniaturization are limited. Here, we demonstrate the α-In2Se3 ferroelectric semiconductor channel device that integrates non-volatile memory (NVM) and neural computation functions. Remarkable performance includes NVM ultra-fast write speed of 40ns, improved endurance through internal electric field, flexible adjustment of neural plasticity, energy consumption as low as 40fJ per event, and thermally modulated 94.74% high-precision iris recognition classification simulation. This prototypical demonstration laid the foundation for an integrated memory computing system with high density and energy efficiency.

scholarly journals Two-dimensional ferroelectric channel transistors integrating ultra-fast memory and neural computing

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Shuiyuan Wang ◽  
Lan Liu ◽  
Lurong Gan ◽  
Huawei Chen ◽  
Xiang Hou ◽  
...  
Keyword(s):  
Energy Efficiency ◽  
Neural Plasticity ◽  
Iris Recognition ◽  
Computing System ◽  
Internal Electric Field ◽  
Fast Switching ◽  
Analog Computing ◽  
Non Volatile Memory ◽  
Non Destructive ◽  
With Memory

AbstractWith the advent of the big data era, applications are more data-centric and energy efficiency issues caused by frequent data interactions, due to the physical separation of memory and computing, will become increasingly severe. Emerging technologies have been proposed to perform analog computing with memory to address the dilemma. Ferroelectric memory has become a promising technology due to field-driven fast switching and non-destructive readout, but endurance and miniaturization are limited. Here, we demonstrate the α-In2Se3 ferroelectric semiconductor channel device that integrates non-volatile memory and neural computation functions. Remarkable performance includes ultra-fast write speed of 40 ns, improved endurance through the internal electric field, flexible adjustment of neural plasticity, ultra-low energy consumption of 234/40 fJ per event for excitation/inhibition, and thermally modulated 94.74% high-precision iris recognition classification simulation. This prototypical demonstration lays the foundation for an integrated memory computing system with high density and energy efficiency.

Energy efficiency for cloud computing system based on predictive optimization

2017 ◽  
Vol 102 ◽  
pp. 103-114 ◽  
Author(s):  
Dinh-Mao Bui ◽  
YongIk Yoon ◽  
Eui-Nam Huh ◽  
SungIk Jun ◽  
Sungyoung Lee
Keyword(s):  
Energy Efficiency ◽  
Cloud Computing ◽  
Computing System ◽  
Cloud Computing System

scholarly journals Coordinated Plasticity between Barrel Cortical Glutamatergic and GABAergic Neurons during Associative Memory

2016 ◽  
Vol 2016 ◽  
pp. 1-20 ◽  
Author(s):  
Fenxia Yan ◽  
Zilong Gao ◽  
Pin Chen ◽  
Li Huang ◽  
Dangui Wang ◽  
...  
Keyword(s):  
Associative Memory ◽  
Neural Plasticity ◽  
High Throughput Sequencing ◽  
Memory Formation ◽  
Gabaergic Neurons ◽  
Glutamatergic Neurons ◽  
Sensory Cortices ◽  
With Memory ◽  
Odor Signal ◽  
Epigenetic Processes

Neural plasticity is associated with memory formation. The coordinated refinement and interaction between cortical glutamatergic and GABAergic neurons remain elusive in associative memory, which we examine in a mouse model of associative learning. In the mice that show odorant-induced whisker motion after pairing whisker and odor stimulations, the barrel cortical glutamatergic and GABAergic neurons are recruited to encode the newly learnt odor signal alongside the innate whisker signal. These glutamatergic neurons are functionally upregulated, and GABAergic neurons are refined in a homeostatic manner. The mutual innervations between these glutamatergic and GABAergic neurons are upregulated. The analyses by high throughput sequencing show that certain microRNAs related to regulating synapses and neurons are involved in this cross-modal reflex. Thus, the coactivation of the sensory cortices through epigenetic processes recruits their glutamatergic and GABAergic neurons to be the associative memory cells as well as drive their coordinated refinements toward the optimal state for the storage of the associated signals.

scholarly journals Lightweight hardware fingerprinting solution using inherent memory in off-the-shelf commodity devices

2022 ◽  
Vol 25 (1) ◽  
pp. 105
Author(s):  
Mohd Syafiq Mispan ◽  
Aiman Zakwan Jidin ◽  
Muhammad Raihaan Kamarudin ◽  
Haslinah Mohd Nasir
Keyword(s):  
Integrated Circuit ◽  
Trusted Computing ◽  
Random Access ◽  
Process Variations ◽  
Computing System ◽  
Access Memory ◽  
Start Up ◽  
Non Volatile Memory ◽  
Memory Power ◽  
On Chip

An emerging technology known as Physical unclonable function (PUF) can provide a hardware root-of-trust in building the trusted computing system. PUF exploits the intrinsic process variations during the integrated circuit (IC) fabrication to generate a unique response. This unique response differs from one PUF to the other similar type of PUFs. Static random-access memory PUF (SRAM-PUF) is one of the memory-based PUFs in which the response is generated during the memory power-up process. Non-volatile memory (NVM) architecture like SRAM is available in off-the-shelf microcontroller devices. Exploiting the inherent SRAM as PUF could wide-spread the adoption of PUF. Therefore, in this study, we evaluate the suitability of inherent SRAM available in ATMega2560 microcontroller on Arduino platform as PUF that can provide a unique fingerprint. First, we analyze the start-up values (SUVs) of memory cells and select only the cells that show random values after the power-up process. Subsequently, we statistically analyze the characteristic of fifteen SRAM-PUFs which include uniqueness, reliability, and uniformity. Based on our findings, the SUVs of fifteen on-chip SRAMs achieve 42.64% uniqueness, 97.28% reliability, and 69.16% uniformity. Therefore, we concluded that the available SRAM in off-the-shelf commodity hardware has good quality to be used as PUF.

scholarly journals Measuring And Computing System for Studying the Magnetic Characteristics of Soft Magnetic Materials

2021 ◽  
pp. 21-26
Author(s):  
A.V. Egorov ◽  
V.V. Polyakov ◽  
A.A. Lependin ◽  
D.D. Ruder
Keyword(s):  
Magnetic Properties ◽  
Eddy Current ◽  
Magnetic Permeability ◽  
Computing System ◽  
Ferromagnetic Materials ◽  
Soft Magnetic Materials ◽  
Magnetic Characteristics ◽  
Soft Magnetic ◽  
Non Destructive

Non-destructive eddy current diagnostics of the structure, composition, physical and mechanical properties of ferromagnetic materials, as well as eddy current monitoring of the operational parameters of products manufactured from them, requires knowledge of the magnetic characteristics of these materials. In eddy current measurements, the results obtained are influenced by a significant number of factors — magnetic and electrical properties of materials, geometric characteristics of products, measurement conditions, design features of an eddy current sensor, etc. Also, the magnetic properties themselves have high structural sensitivity. Thus, identification of the diagnosed parameters puts great importance on the tasks to separate the influencing factors and isolate the contribution of the magnetic properties. This paper describes the measuring and computing system that allows automatic determination of the magnetic permeability of soft magnetic ferromagnetic materials at various values of the strength of the external magnetizing field. The system has been tested using soft magnetic ferrites samples. An experimental dependence of the magnetic permeability on the magnitude of the magnetic field for the initial section of the main magnetization curve is presented. The obtained initial magnetic permeability is compared with the data of independent indirect measurements. The proposed system provides an increase in the reliability and accuracy of the results of the experimental determination of magnetic characteristics and can be used for non-destructive diagnostics of products made of soft magnetic ferromagnetic materials.

scholarly journals Resonance for analog recurrent neural network

2021 ◽  
Author(s):  
Yurui Qu ◽  
Ming Zhou ◽  
Erfan Khoram ◽  
Nanfang Yu ◽  
Zongfu Yu
Keyword(s):  
Recurrent Neural Networks ◽  
Computing System ◽  
Long Term Memory ◽  
Neural Computing ◽  
Short Term ◽  
Term Memory ◽  
Fast Speed ◽  
Analog Computing ◽  
Intrinsic Parallelism

Abstract There is a strong interest in using physical waves for artificial neural computing because of their unique advantages in fast speed and intrinsic parallelism. Resonance, as a ubiquitous feature across many wave systems, is a natural candidate for analog computing in temporal signals. We demonstrate that resonance can be used to construct stable and scalable recurrent neural networks. By including resonators with different lifetimes, the computing system develops both short-term and long-term memory simultaneously.

scholarly journals Programmable terahertz metamaterials with non-volatile memory

2021 ◽  
Author(s):  
Benwen Chen ◽  
Jingbo Wu ◽  
Weili Li ◽  
Caihong Zhang ◽  
Qiang Xue ◽  
...  
Keyword(s):  
Memory Effect ◽  
Large Scale ◽  
Spatial Light Modulators ◽  
Switching Speed ◽  
Compressive Imaging ◽  
Light Modulators ◽  
Non Volatile Memory ◽  
Multi Level ◽  
Change Material ◽  
With Memory

Abstract Spatial light modulators (SLMs) exhibited the powerful capability of controlling the electromagnetic wave. They have found numerous applications at terahertz (THz) frequencies, including wireless communication, digital holography, and compressive imaging. However, the development towards large-scale, multi-level and multi-functional THz SLM encounters technical challenges. Here, we present an electrically programmable THz metamaterial consisting of an array of 8⊆8 pixels, in which the phase change material of vanadium dioxide (VO2) is embedded. After successfully suppressing the crosstalk from adjacent pixels, the THz wave could be modulated in a programmable manner. The switching speed of each pixel was on the order of 1 kHz. In particular, utilising the hysteresis effect of VO2, the memory effect is demonstrated. The THz amplitude of each pixel can be written and erased by individual current pulses. Furthermore, multi-state THz images could be generated and stored, with a retention time of more than 5 hours. This programmable metamaterial with memory effect can be extended to other frequency bands and opens a route for electromagnetic information processing.

Sign in / Sign up

Export Citation Format

Share Document