ieeexplore最新文献

Blank Page 空白页

IF 1.3 4区物理与天体物理

IEEE Transactions on Plasma Science Pub Date : 2025-01-17 DOI: 10.1109/TPS.2025.3526510

引用次数: 0

2024 Index IEEE Transactions on Device and Materials Reliability Vol. 24 器件与材料可靠性学报，第24卷

IF 2.5 3区工程技术

IEEE Transactions on Device and Materials Reliability Pub Date : 2025-01-14 DOI: 10.1109/TDMR.2025.3528093

引用次数: 0

IEEE Sensors Council IEEE传感器委员会

IF 4.3 2区综合性期刊

IEEE Sensors Journal Pub Date : 2025-01-13 DOI: 10.1109/JSEN.2024.3525383

引用次数: 0

IEEE Open Journal of Nanotechnology Information for Authors IEEE纳米技术信息开放杂志作者

IF 1.8

IEEE Open Journal of Nanotechnology Pub Date : 2025-01-06 DOI: 10.1109/OJNANO.2025.3525915

引用次数: 0

IEEE Sensors Council IEEE传感器委员会

IF 4.3 2区综合性期刊

IEEE Sensors Journal Pub Date : 2025-01-03 DOI: 10.1109/JSEN.2024.3520066

引用次数: 0

Approximation-Aware Training for Efficient Neural Network Inference on MRAM Based CiM Architecture 基于MRAM的CiM结构中高效神经网络推理的逼近感知训练

IF 1.8

IEEE Open Journal of Nanotechnology Pub Date : 2024-12-31 DOI: 10.1109/OJNANO.2024.3524265

Hemkant Nehete;Sandeep Soni;Tharun Kumar Reddy Bollu;Balasubramanian Raman;Brajesh Kumar Kaushik

{"title":"Approximation-Aware Training for Efficient Neural Network Inference on MRAM Based CiM Architecture","authors":"Hemkant Nehete;Sandeep Soni;Tharun Kumar Reddy Bollu;Balasubramanian Raman;Brajesh Kumar Kaushik","doi":"10.1109/OJNANO.2024.3524265","DOIUrl":"https://doi.org/10.1109/OJNANO.2024.3524265","url":null,"abstract":"Convolutional neural networks (CNNs), despite their broad applications, are constrained by high computational and memory requirements. Existing compression techniques often neglect approximation errors incurred during training. This work proposes approximation-aware-training, in which group of weights are approximated using a differential approximation function, resulting in a new weight matrix composed of approximation function's coefficients (AFC). The network is trained using backpropagation to minimize the loss function with respect to AFC matrix with linear and quadratic approximation functions preserving accuracy at high compression rates. This work extends to implement an compute-in-memory architecture for inference operations of approximate neural networks. This architecture includes a mapping algorithm that modulates inputs and map AFC to crossbar arrays directly, eliminating the need to predict approximated weights for evaluating output. This reduces the number of crossbars, lowering area and energy consumption. Integrating magnetic random-access memory-based devices further enhances performance by reducing latency and energy consumption. Simulation results on approximated LeNet-5, VGG8, AlexNet, and ResNet18 models trained on the CIFAR-100 dataset showed reductions of 54%, 30%, 67%, and 20% in the total number of crossbars, respectively, resulting in improved area efficiency. In the ResNet18 architecture, latency and energy consumption decreased by 95% and 93.3% with spin-orbit torque (SOT) based crossbars compared to RRAM-based architectures.","PeriodicalId":446,"journal":{"name":"IEEE Open Journal of Nanotechnology","volume":"6 ","pages":"16-26"},"PeriodicalIF":1.8,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10819260","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142993286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improvement of Surface Roughness in SiO2 Thin Films via Deuterium Annealing at 300 °C 300℃氘退火改善SiO2薄膜表面粗糙度的研究

IF 2.1 4区工程技术

IEEE Transactions on Nanotechnology Pub Date : 2024-12-31 DOI: 10.1109/TNANO.2024.3524567

Ju-Won Yeon;Hyo-Jun Park;Eui-Cheol Yun;Moon-Kwon Lee;Tae-Hyun Kil;Yong-Sik Kim;Jun-Young Park

引用次数: 0

On the Importance of the Metal Catalyst Layer to the Performance of CNT-Based Supercapacitor Electrodes 金属催化剂层对碳纳米管超级电容器电极性能的重要性

IF 2.1 4区工程技术

IEEE Transactions on Nanotechnology Pub Date : 2024-12-26 DOI: 10.1109/TNANO.2024.3523412

Kingshuk Chatterjee;Vinay Kumar;Prabhat Kumar Agnihotri;Sumit Basu;Nandini Gupta

{"title":"On the Importance of the Metal Catalyst Layer to the Performance of CNT-Based Supercapacitor Electrodes","authors":"Kingshuk Chatterjee;Vinay Kumar;Prabhat Kumar Agnihotri;Sumit Basu;Nandini Gupta","doi":"10.1109/TNANO.2024.3523412","DOIUrl":"https://doi.org/10.1109/TNANO.2024.3523412","url":null,"abstract":"The power and energy densities of a Supercapacitor (SC) is largely dictated by the accessibility of the nano-porous area of the electrode to the electrolyte ions. Carbon nanotubes (CNT) have high electrical conductivity, and more importantly, may be grown into architectures with high surface area. However, this is not easy to achieve in practice. CNT electrodes are fabricated by chemical vapor deposition (CVD), after a metal catalyst layer is coated on a current collector. In this work, the control of the metal catalyst layer, by varying the dip-coating time and CVD process parameters, is shown to be crucial to pore morphology and consequent SC performance. The dip-coating time is adjusted to obtain thin and uniform coating. Further, optimum reduction of the nickel layer with hydrogen is required to produce thin CNTs with adequate inter-tube separation that facilitate ion accessibility within the pores. The height of the CNT forest is also optimized to prevent decrease in specific capacitance due to reduced accessibility. Proper optimization of the process parameters results in a pore morphology conductive to ion diffusion, and simultaneous improvement in energy and power density.","PeriodicalId":449,"journal":{"name":"IEEE Transactions on Nanotechnology","volume":"24 ","pages":"48-53"},"PeriodicalIF":2.1,"publicationDate":"2024-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142940840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Improvement of the Enhancement-Mode GaN MIS-HEMTs by Fluorine Doping in the Dielectric Gate Stack 介质栅层中氟掺杂对增强模式GaN mishemt的改进

IF 2.1 4区工程技术

IEEE Transactions on Nanotechnology Pub Date : 2024-12-25 DOI: 10.1109/TNANO.2024.3522371

Tsung-Ying Yang;Mei-Yan Kuo;Jui-Sheng Wu;Yan-Kui Liang;Rahul Rai;Shivendra K. Rathaur;Edward Yi Chang

{"title":"Improvement of the Enhancement-Mode GaN MIS-HEMTs by Fluorine Doping in the Dielectric Gate Stack","authors":"Tsung-Ying Yang;Mei-Yan Kuo;Jui-Sheng Wu;Yan-Kui Liang;Rahul Rai;Shivendra K. Rathaur;Edward Yi Chang","doi":"10.1109/TNANO.2024.3522371","DOIUrl":"https://doi.org/10.1109/TNANO.2024.3522371","url":null,"abstract":"This study tested fluorine doping on various regions of the ferroelectric charge trap gate stack (FEG stack). Fluorine doping effectively reduces oxygen vacancies in the dielectric layer, thus reducing leakage current and stabilizing charge in the dielectric layer. Moreover, fluorine doping can passivate the dangling bonds at the interface and increase the ability of trapping carriers in the trap layer. The FEG stack comprises a tunnel oxide layer (TL), a charge trap layer (CTL), and a ferroelectric layer (FE). Four types of devices were fabricated: undoped, doping in TL, doping in CTL, and doping in both TL and CTL, to investigate the impact of fluorine doping on the FEG gate stack. Devices doping in TL and CTL demonstrated superior performance, achieving the highest V\u0000<sub>th</sub>\u0000 of 5.4 V with a retention time of 70.42% after 10, 000 seconds. The off-state and gate leakage tests revealed impressive breakdown voltages of 735 V and 24.55 V, respectively. Furthermore, the device exhibited a high operation voltage of 14.3 V for a 10-year lifetime prediction, enabling a wide operating range.","PeriodicalId":449,"journal":{"name":"IEEE Transactions on Nanotechnology","volume":"24 ","pages":"42-47"},"PeriodicalIF":2.1,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142938491","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Member ad suite 会员广告套件

IF 1.3 4区物理与天体物理

IEEE Transactions on Plasma Science Pub Date : 2024-12-25 DOI: 10.1109/TPS.2024.3519497

引用次数: 0