Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963745
Jodh S. Pannu, Sunny Raj, S. Fernandes, Sumit Kumar Jha, Dwaipayan Chakraborty, Sarah Rafiq, N. Cady
Detection of edges in images is an elementary operation in computer vision that can greatly benefit from an implementation with a low power-delay product. In this paper, we propose a new approach for designing nanoscale memristor crossbars that can implement approximate edge-detection using flow-based computing. Instead of the traditional Boolean approach, our methodology uses a ternary logic approach with three outcomes: True representing an edge, False that representing the absence of an edge, and Don’t Care that represents an ambivalent response. Our data-driven design approach uses a corpus of human-labeled edges in order to learn the concept of an edge in an image. A massively parallel simulated annealing search algorithm over 96 processes is used to obtain the design of the memristor crossbar for edge detection. We show that our approximate crossbar design is effective in computing edges of images on the BSD500 benchmark.
{"title":"Data-driven Approximate Edge Detection using Flow-based Computing on Memristor Crossbars","authors":"Jodh S. Pannu, Sunny Raj, S. Fernandes, Sumit Kumar Jha, Dwaipayan Chakraborty, Sarah Rafiq, N. Cady","doi":"10.1109/ANS47466.2019.8963745","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963745","url":null,"abstract":"Detection of edges in images is an elementary operation in computer vision that can greatly benefit from an implementation with a low power-delay product. In this paper, we propose a new approach for designing nanoscale memristor crossbars that can implement approximate edge-detection using flow-based computing. Instead of the traditional Boolean approach, our methodology uses a ternary logic approach with three outcomes: True representing an edge, False that representing the absence of an edge, and Don’t Care that represents an ambivalent response. Our data-driven design approach uses a corpus of human-labeled edges in order to learn the concept of an edge in an image. A massively parallel simulated annealing search algorithm over 96 processes is used to obtain the design of the memristor crossbar for edge detection. We show that our approximate crossbar design is effective in computing edges of images on the BSD500 benchmark.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122556603","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963738
Mami Kubota, K. Takanashi, K. Dunn
Artificial Intelligence techniques (AI) offer the possibility of stabilize process control independent of human skills for many industries, including Chemical Mechanical Planarization (CMP) in a silicon wafer manufacturing flow. In this arena, AI is mainly used to develop a formula for predicting the quality of wafers after CMP. In the current stage of AI development, explanatory variables are required to be quantifiable values, and still need to be chosen by a human. Generally, the explanatory variables used are the processing data which is easy to get from CMP equipment in a manufacturing process. However, these values are not directly measuring the conditions of the wafer or the consumables. In this study, we focus on providing more directly relevant data which could serve as inputs for AI algorithms based on the condition of the silica particles in the CMP slurry. In particular, we quantify the agglomeration levels of silica particles (AGL), and investigate the behaviours of AGL at several pressure levels. Further, we explain the relationships between AGL and polishing abilities, as a more relevant input for AI prediction of wafer quality.
{"title":"Investigation of Quantification of Agglomeration Level of Silica Particles in CMP Slurry for Creating the Quality Prediction Formula by AI Technology","authors":"Mami Kubota, K. Takanashi, K. Dunn","doi":"10.1109/ANS47466.2019.8963738","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963738","url":null,"abstract":"Artificial Intelligence techniques (AI) offer the possibility of stabilize process control independent of human skills for many industries, including Chemical Mechanical Planarization (CMP) in a silicon wafer manufacturing flow. In this arena, AI is mainly used to develop a formula for predicting the quality of wafers after CMP. In the current stage of AI development, explanatory variables are required to be quantifiable values, and still need to be chosen by a human. Generally, the explanatory variables used are the processing data which is easy to get from CMP equipment in a manufacturing process. However, these values are not directly measuring the conditions of the wafer or the consumables. In this study, we focus on providing more directly relevant data which could serve as inputs for AI algorithms based on the condition of the silica particles in the CMP slurry. In particular, we quantify the agglomeration levels of silica particles (AGL), and investigate the behaviours of AGL at several pressure levels. Further, we explain the relationships between AGL and polishing abilities, as a more relevant input for AI prediction of wafer quality.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117313580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963740
K. Petrillo, D. Hetzer, Takashi Shimoaoki, Yusaku Hashimoto, Kouichirou Tanaka, S. Kawakami, Cody Murray, Luciana Meli, A. Hubbard, Saumya Sharma, A. de Silva, L. Huli, N. Shibata, Corey Lemley
With the insertion of EUV lithography into high volume manufacturing, mature lithographic materials and processes are required on multiple fronts. Not only do lithographic materials require optimization to reduce stochastic effects; processing techniques, underlying films, and auxiliary processes such as rinse are also known to have important impacts on the ability to yield sub-36nm pitch devices. In this paper we will describe the contribution that resist maturity, improved underlayers, rinse materials, and process parameters such as develop optimization and improved hardware have on yield improvement.
{"title":"Material and process improvements towards sub 36nm pitch EUV single exposure","authors":"K. Petrillo, D. Hetzer, Takashi Shimoaoki, Yusaku Hashimoto, Kouichirou Tanaka, S. Kawakami, Cody Murray, Luciana Meli, A. Hubbard, Saumya Sharma, A. de Silva, L. Huli, N. Shibata, Corey Lemley","doi":"10.1109/ANS47466.2019.8963740","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963740","url":null,"abstract":"With the insertion of EUV lithography into high volume manufacturing, mature lithographic materials and processes are required on multiple fronts. Not only do lithographic materials require optimization to reduce stochastic effects; processing techniques, underlying films, and auxiliary processes such as rinse are also known to have important impacts on the ability to yield sub-36nm pitch devices. In this paper we will describe the contribution that resist maturity, improved underlayers, rinse materials, and process parameters such as develop optimization and improved hardware have on yield improvement.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114873780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963742
E. C. Graham, Nicholas M. Fahrenkopf, N. Cady
To study how though- oxide-via (TOV) induced stress affects light propagation, a numerical model is constructed using the commercially available finite element analysis (FEA) software COMSOL Multiphysics 5.4a. In the model, directional couplers are used to analyze wavelength shifts as TOV structures are placed at varying distances away. Increasing the number of TOVs in proximity to directional couplers is also explored. Finally, determining the need for a keep-out-zone is assesed for directional couplers fabricated in a silicon photonic interposer.
{"title":"Through Oxide Via (TOV) Induced Fabrication Stress on Directional Couplers in a Si Photonic Interposer","authors":"E. C. Graham, Nicholas M. Fahrenkopf, N. Cady","doi":"10.1109/ANS47466.2019.8963742","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963742","url":null,"abstract":"To study how though- oxide-via (TOV) induced stress affects light propagation, a numerical model is constructed using the commercially available finite element analysis (FEA) software COMSOL Multiphysics 5.4a. In the model, directional couplers are used to analyze wavelength shifts as TOV structures are placed at varying distances away. Increasing the number of TOVs in proximity to directional couplers is also explored. Finally, determining the need for a keep-out-zone is assesed for directional couplers fabricated in a silicon photonic interposer.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126871506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963739
H. Landis, Gazi Huda, J. Sucharitaves
Using swap/unit cells in FEOL Fill is a widely practiced method because of the placement algorithm simplicity. Over the last several years, Fill engineers devised intelligent ways to find multiple optimum sized swap cells and have been placing them sequentially to achieve higher Fill density. In cutting edge FinFET technologies nodes, however, such swap cell approach cannot meet the aggressive process requirements, such as demanding density, and due to inter-layer dependencies in FEOL swap cell.To overcome this challenge, GLOBALFOUNDRIES (GF) Fill algorithm has removed dummy Fin from swap cells to place at the very beginning of FEOL Fill routine. The subsequent cells align with previously placed "Fin-grid". Such adaptation ensures meeting the process requirements of surrounding FEOL Fill layers by RxFin, while a higher FEOL Fill density is achieved with a high efficiency. Along this line, the associated implementation choices and the corresponding tradeoffs are discussed.
{"title":"Gridded Placement of Integrated Unit Cells for FEOL Fill","authors":"H. Landis, Gazi Huda, J. Sucharitaves","doi":"10.1109/ANS47466.2019.8963739","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963739","url":null,"abstract":"Using swap/unit cells in FEOL Fill is a widely practiced method because of the placement algorithm simplicity. Over the last several years, Fill engineers devised intelligent ways to find multiple optimum sized swap cells and have been placing them sequentially to achieve higher Fill density. In cutting edge FinFET technologies nodes, however, such swap cell approach cannot meet the aggressive process requirements, such as demanding density, and due to inter-layer dependencies in FEOL swap cell.To overcome this challenge, GLOBALFOUNDRIES (GF) Fill algorithm has removed dummy Fin from swap cells to place at the very beginning of FEOL Fill routine. The subsequent cells align with previously placed \"Fin-grid\". Such adaptation ensures meeting the process requirements of surrounding FEOL Fill layers by RxFin, while a higher FEOL Fill density is achieved with a high efficiency. Along this line, the associated implementation choices and the corresponding tradeoffs are discussed.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127918162","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963743
H. Nieto-Chaupis
We use the Quantum Mechanics path integral to describe bacteria spatial dynamics inside a box that that is limited by constant electric potentials. While this potential can be materialized to some extent of being a physical representation of macro-phage cell the bacteria behavior is also driven by the Jackson’s electric potential. In this manner, the resultant repulsive or attractive electric force might drive the spatial displacement of bacteria to accomplish the task of motility and chemotaxis. The path-integral language seen a stochastic tool appears as an interesting option to model bacteria and Antimicrobial peptides agents when both of them are in a random struggle in space-time. In addition its complementary to Electrodynamics support tits usage inside of the Microbiological territory. Our hybrid approach turns out to be precise with an error of order of 3%.
{"title":"Nanodevices Versus Bacteria in a Box: The Correspondence between Classical Electrodynamics and the Quantum Mechanics Path Integral","authors":"H. Nieto-Chaupis","doi":"10.1109/ANS47466.2019.8963743","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963743","url":null,"abstract":"We use the Quantum Mechanics path integral to describe bacteria spatial dynamics inside a box that that is limited by constant electric potentials. While this potential can be materialized to some extent of being a physical representation of macro-phage cell the bacteria behavior is also driven by the Jackson’s electric potential. In this manner, the resultant repulsive or attractive electric force might drive the spatial displacement of bacteria to accomplish the task of motility and chemotaxis. The path-integral language seen a stochastic tool appears as an interesting option to model bacteria and Antimicrobial peptides agents when both of them are in a random struggle in space-time. In addition its complementary to Electrodynamics support tits usage inside of the Microbiological territory. Our hybrid approach turns out to be precise with an error of order of 3%.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117236859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ANS47466.2019.8963741
A. Chandra, P. Oldiges, Ching-Tzu Chen, T. Philip, P. Adusumilli, M. BrightSky
The bottom electrode (heater) in a mushroom cell phase change memory (PCM) is one of the major components which determines device performance. Understanding the effect of heater geometry on Reset characteristics is necessary for engineering and fabricating efficient structures. In this paper, we explore the role of certain geometrical parameters of the heater using computer simulations. Performing such parameter sweeps experimentally would be uneconomical and time consuming. Our results indicate that the heater top radius/area is of utmost importance for Reset Current and cross-sectional area of the heater matters more that the actual shape of the cross-section. The focus of this paper is to provide a qualitative understanding of the variation of quantities of interest as the heater configuration is changed.
{"title":"Impact of heater configuration on Reset characteristics of PCM Mushroom cell","authors":"A. Chandra, P. Oldiges, Ching-Tzu Chen, T. Philip, P. Adusumilli, M. BrightSky","doi":"10.1109/ANS47466.2019.8963741","DOIUrl":"https://doi.org/10.1109/ANS47466.2019.8963741","url":null,"abstract":"The bottom electrode (heater) in a mushroom cell phase change memory (PCM) is one of the major components which determines device performance. Understanding the effect of heater geometry on Reset characteristics is necessary for engineering and fabricating efficient structures. In this paper, we explore the role of certain geometrical parameters of the heater using computer simulations. Performing such parameter sweeps experimentally would be uneconomical and time consuming. Our results indicate that the heater top radius/area is of utmost importance for Reset Current and cross-sectional area of the heater matters more that the actual shape of the cross-section. The focus of this paper is to provide a qualitative understanding of the variation of quantities of interest as the heater configuration is changed.","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127987292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ans47466.2019.8963744
N. Lanzillo, R. Robison
We demonstrate a simulation workflow based on first-principles calculations to rapidly screen candidate materials for viability as ferromagnetic electrodes in magnetic tunnel junctions (MTJs) for the next generation of high-performance magnetic random access memory (MRAM) technology. For a series of Fe-based alloys with a fixed crystal structure, we calculate formation energies, bulk spin polarization, and essential magnetic properties including magnetic anisotropy energy (MAE) and tunneling magnetoresistance (TMR). This work demonstrates a materials optimization strategy that can guide on-wafer experiments
{"title":"A Materials Screening Methodology for Scaled Non-Volatile Memory in the AI Era","authors":"N. Lanzillo, R. Robison","doi":"10.1109/ans47466.2019.8963744","DOIUrl":"https://doi.org/10.1109/ans47466.2019.8963744","url":null,"abstract":"We demonstrate a simulation workflow based on first-principles calculations to rapidly screen candidate materials for viability as ferromagnetic electrodes in magnetic tunnel junctions (MTJs) for the next generation of high-performance magnetic random access memory (MRAM) technology. For a series of Fe-based alloys with a fixed crystal structure, we calculate formation energies, bulk spin polarization, and essential magnetic properties including magnetic anisotropy energy (MAE) and tunneling magnetoresistance (TMR). This work demonstrates a materials optimization strategy that can guide on-wafer experiments","PeriodicalId":375888,"journal":{"name":"2019 IEEE Albany Nanotechnology Symposium (ANS)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117183602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：