检索结果-内蒙古大学图书馆

Parallelization of Lattice Boltzmann method using MPI domain decomposition technology for a drop impact on a wetted solid wall

引用

International Journal of Modeling, Simulation, and Scientific computing 2014年第2期5卷 40-52页

作者： Zhi Shang Ming Cheng Jing Lou Institute of High Performance Computing Agency for Science Technology and Research(A*STAR)1 Fusionopolis Way#16-16 ConnexisSingapore 138632

Lattice Boltzmann method(LBM)is a new attractive computational approach for simulating isothermal multi-phase flows in computational fluid dynamics(CFD).It is based on the kinetic theory and easy to be *** study aims to analyze the performance of parallel LBM programming for the incompressible two-phase flows at high density and viscosity *** this purpose,a liquid drop impact on a wetted wall with a pre-existing thin film of the same liquid is simulated by using the parallel LBM *** the simulations,the domain decomposition,data communication and parallelization of the LBM code using the message passing interface(MPI)library have been *** computational results show that the parallel LBM code exhibits a good high performance computing(HPC)on the parallel speed-up.

关键词： Dropimpact two-phaseflow Lattice Boltzmann method MPI domain decomposition HPC

来源：评论

学校读者我要写书评

暂无评论

A comprehensive study in assembling deep convolutional neural networks for image classification 10

A comprehensive study in assembling deep convolutional neura...

引用

10th International Conference on Signal Processing Systems, ICSPS 2018

作者： Rom, Uddamvathanak Yang, Feng School of Engineering and Technology PSB Academy Singapore Department of Computing Science Institute of High Performance Computing ASTAR Singapore

ISBN: (数字)9781510628366

ISBN: (纸本)9781510628359

Ensemble methods have been broadly used in many applications to integrate individual models for achieving better performance like accuracy and robustness. This paper focuses on using deep Convolutional Neural Networks (CNNs) models and ensembling them for image classification. To have a comprehensive and comparative study, five ensemble techniques are explored and their combinations were applied in a two-stage ensemble process. In detail, multiple basic CNN models (structures) are first pre-defined to increase model variety for the ensemble, from which each basic CNN model is trained multiple rounds based on sub-sampling on the training dataset. In the first stage of ensemble, multiple predictions (through sub-sampling) from each basic CNN structure are combined. This is followed by the second stage of ensemble to integrate the outputs from all basic CNN structures. Experiments were conducted using Kaggle's 'Statoil/CCORE Iceberg Classifier Challenge' image data for iceberg and ship classification. The experimental results showed that the ensembling CNN models could improve classification accuracy in the image classification problem. © 2019 SPIE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Optimizing and auto-tuning scale-free sparse matrix-vector multiplication on Intel Xeon Phi

Optimizing and auto-tuning scale-free sparse matrix-vector m...

引用

2015 IEEE/ACM International Symposium on Code Generation and Optimization, CGO 2015

作者： Tang, Wai Teng Zhao, Ruizhe Lu, Mian Liang, Yun Huyng, Huynh Phung Li, Xibai Goh, Rick Siow Mong Institute of High Performance Computing Agency for Science Technology and Research Singapore Center for Energy-Efficient Computing and Applications School of EECS Peking University China

ISBN: (纸本)9781479981618

Recently, the Intel Xeon Phi coprocessor has received increasing attention in high performance computing due to its simple programming model and highly parallel architecture. In this paper, we implement sparse matrix vector multiplication (SpMV) for scale-free matrices on the Xeon Phi architecture and optimize its performance. Scale-free sparse matrices are widely used in various application domains, such as in the study of social networks, gene networks and web graphs. We propose a novel SpMV format called vectorized hybrid COO+CSR (VHCC). Our SpMV implementation employs 2D jagged partitioning, tiling and vectorized prefix sum computations to improve hardware resource utilization, and thus overall performance. As the achieved performance depends on the number of vertical panels, we also develop a performance tuning method to guide its selection. Experimental results demonstrate that our SpMV implementation achieves an average 3× speedup over Intel MKL for a wide range of scale-free matrices. © 2015 IEEE.

关键词： Parallel architectures

来源：评论

学校读者我要写书评

暂无评论

A Digital 3D TCAM Accelerator for the Inference Phase of Random Forest 23

A Digital 3D TCAM Accelerator for the Inference Phase of Ran...

引用

Proceedings of the 60th Annual ACM/IEEE Design Automation Conference

作者： Chieh-Lin Tsai Chun-Feng Wu Yuan-Hao Chang Han-Wen Hu Yung-Chun Lee Hsiang-Pang Li Tei-Wei Kuo Department of Computer Science and Information Engineering National Taiwan University Department of Computer Science National Yang Ming Chiao Tung University Institute of Information Science Academia Sinica Taiwan Macronix International Co. Ltd and Department of Electrical Engineering National Tsing Hua University Macronix International Co. Ltd Department of Computer Science and Information Engineering National Taiwan University and Mohamed Bin Zayed University of Artificial Intelligence and NTU High Performance and Scientific Computing Center National Taiwan University

ISBN: (纸本)9798350323481

Random forest is a popular ensemble machine-learning algorithm for classification and regression tasks. However, the irregular tree shapes and non-deterministic memory access patterns make it hard for the current von Neumann architecture to handle random forest efficiently. This paper proposes a digital 3D TCAM-based accelerator for the random forest, adopting the idea of processing-in-memory (PIM) to reduce data movement. By utilizing this accelerator, we propose a TCAM-based approach to provide real-time inference with low energy consumption, making it suitable for edge or embedded environments. In the experiments, the proposed approach achieves an average of 3.13 times higher throughput with 22 times more energy saving than the GPU approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Feasibility Study on Restoring Chinese Documents by Convolutional Autoencoder

Feasibility Study on Restoring Chinese Documents by Convolut...

引用

2023 International Conference on Consumer Electronics - Taiwan, ICCE-Taiwan 2023

作者： Lo, Shi Wei Chou, Hsiu-Mei National Center for High-Performance Computing Data Science and Technology Division Hsinchu Taiwan

ISBN: (纸本)9798350324174

Digitalizing documents has become a crucial aspect of modern information management. However, various factors, such as manual handwriting, stains, and degradation, can significantly impact the quality of digitalized documents. This paper proposes a autoencoder-based network cleaning technology to completely erase manual writing, stains, colors, and background noise from paper documents to enhance their digitalized quality. Our proposed approach utilizes machine learning algorithms to identify and isolate unwanted elements from the document and then replace them with clean and transparent backgrounds. The experimental results demonstrate that our proposed approach significantly improves the digitalized quality of paper documents, making them more suitable for long-term storage and retrieval. Our proposed technology has the potential to revolutionize the field of paper document digitalization and improve the efficiency of information management. © 2023 IEEE.

关键词： Information management

来源：评论

学校读者我要写书评

暂无评论

Evaluating and Improving Optical Character Recognition (OCR) Efficiency in Recognizing Mandarin Phrases with Phonetic Symbols

Evaluating and Improving Optical Character Recognition (OCR)...

引用

2022 IEEE International Conference on Internet of Things and Intelligence Systems, IoTaIS 2022

作者： Lo, Shi Wei Chou, Hsiu-Mei National Center for High-Performance Computing Data Science and Technology Division Hsinchu Taiwan

ISBN: (数字)9798350396454

ISBN: (纸本)9798350396454

The development of tools and materials for those people who are visually impaired is a crucial topic in the research area of assistive technologies. In mandarin learning, most printing materials for K-4 children have phonetic symbols to assist students in learning pronunciation. One challenge in the research area is recognizing the mandarin phrases in those printing materials and transforming them into audiobooks or Braille. To our knowledge, this is the first study examining the Optical Character Recognition (OCR) performance toward phonetic symbols in mandarin. In this study, we conduct experiments on recognizing images with mandarin phrases with phonetic symbols by the side using the OCR system. We propose candidate methods to improve recognition efficiency in the future based on preliminary results. © 2022 IEEE.

关键词： Optical character recognition

来源：评论

学校读者我要写书评

暂无评论

Search Complexity: A Way for the Quantitative Analysis of the Search Space 14th

Search Complexity: A Way for the Quantitative Analysis of th...

引用

14th International Conference on Algorithmic Aspects in Information and Management, AAIM 2020

作者： Ning, Li Zhang, Yong Center for High Performance Computing Shenzhen Institutes of Advanced Technology CAS Shenzhen China

ISBN: (纸本)9783030576011

We considered the situations when search is performed in the structured spaces. In order to isolate the effort of reducing the uncertainty about the information of the space structure from the implementation cost of processing the acquired knowledge, we introduced the concept of search complexity, which describes the quantity order of the information that is necessary to discover the optimum. Furthermore, the proposed concept has been extended to the cases when an approximation to the optimum is acceptable, and we have investigated a situation, in which a solution with the multiplicative error (Formula Presented) and the additive error (Formula Presented) can be found in a relatively low search complexity. © 2020, Springer Nature Switzerland AG.

关键词： Approximation Search complexity Search space

来源：评论

学校读者我要写书评

暂无评论

Predicting integrated thermal and acoustic performance in naturally ventilated high-rise buildings using CFD and FEM simulation

引用

建筑模拟（英文版） 2018年第3期11卷 507-518页

作者： Xiang Yu Qide Zhang Jian Kang Fangsen Cui Institute of High Performance Computing Agency for Science Technology and Research (A*STAR) Singapore 138632 School of Architecture University of Sheffield Sheffield S102TN UK

The study of ventilation windows for both natural ventilation and noise mitigation has drawn significant attention recently. This paper presents the numerical approaches to analyse the integrated thermal and acoustical performance of ventilation windows, for a residential building in tropical climate which employs double-layer noise mitigation window for natural ventilation. Given a set of outdoor wind conditions, the distributions of indoor flow and temperature fields are simulated using Computational Fluid Dynamics (CFD) model. The thermal comfort is evaluated using statistical Predicted Mean Vote (PMV) method. For the acoustic performance, noise radiation from road traffic is assumed as the noise source, and the sound insulation of building fa?ade is simulated using Finite Element Method (FEM). From the simulation results, it is found that the thermal satisfaction response is closely related to the inlet wind temperature and speed, and the window opening size greatly affects the ventilation performance. From the case study in Singapore, during certain season, day/night time and with sufficient wind flow, the ventilation window can provide enough fresh air, maintain adequate thermal comfort and quiet acoustic environment for the occupants. The numerical approaches presented in this paper are applicable to general window design studies, and the simulation findings can be incorporated into green building planning. The advantages of using simulation approaches are highlighted and their limitations are discussed.

关键词： natural ventilation thermal comfort ventilation window sound insulation

来源：评论

学校读者我要写书评

暂无评论

Accelerate Data Loading in Distributed Training with Fast Reconfigurable Optical Network

Accelerate Data Loading in Distributed Training with Fast Re...

引用

2024 Asia Communications and Photonics Conference, ACP 2024 and International Conference on Information Photonics and Optical Communications, IPOC 2024

作者： Li, Wenzhe Zhang, Xinhang Jia, Wenxuan Chen, Zhuohan Yuan, Guojun Wang, Zhan Zhang, Peiheng High Performance Computing Center Institute of Computing Technology CAS Beijing China Department of Intelligent Science and Technology Hangzhou Institute of Advanced Study UCAS Hangzhou China Institute of Intelligent Computing Technology Chinese Academy of Sciences Suzhou China

We enhance our prior all-optical reconfigurable network to accelerate the data loading with minimized communication distance between compute and storage nodes. A novel corresponding data loading scheme is proposed to ... 详细信息

ISBN: (纸本)9798350379266

关键词： Data Loading Distributed Deep Learning Reconfigurable optical network

来源：评论

学校读者我要写书评

暂无评论

Reconstructing the dark energy density in light of DESI BAO observations

arXiv

引用

arXiv 2025年

作者： Berti, Maria Bellini, Emilio Bonvin, Camille Kunz, Martin Viel, Matteo Zumalacarregui, Miguel Département de Physique Théorique Center for Astroparticle Physics Université de Genève Quai E. Ansermet 24 Genève 4CH-1211 Switzerland Center for Astrophysics and Cosmology University of Nova Gorica Nova Gorica Slovenia INAF Istituto Nazionale di Astrofisica Osservatorio Astronomico di Trieste Via G. B. Tiepolo 11 TriesteI-34131 Italy IFPU Institute for Fundamental Physics of the Universe via Beirut 2 TriesteI-34151 Italy SISSA - International School for Advanced Studies Via Bonomea 265 TriesteI-34136 Italy INFN – National Institute for Nuclear Physics Via Valerio 2 TriesteI-34127 Italy Italian Research Center on High Performance Computing Big Data and Quantum Computing Italy Am Mühlenberg 1 Potsdam-GolmD-14476 Germany

In light of the evidence for dynamical dark energy (DE) found from the most recent Dark Energy Spectroscopic Instrument (DESI) baryon acoustic oscillation (BAO) measurements, we perform a non-parametric, model-independent reconstruction of the DE density evolution. To do so, we develop and validate a new framework that reconstructs the DE density through a third-degree piece-wise polynomial interpolation, allowing for direct constraints on its redshift evolution without assuming any specific functional form. The strength of our approach resides in the choice of directly reconstructing the DE density, which provides a more straightforward relation to the distances measured by BAO than the equation of state parameter. We investigate the constraining power of cosmic microwave background (CMB) observations combined with supernovae (SNe) and BAO measurements. In agreement with results from other works, we find a preference for deviations from ΛCDM, with a significance of 2.4σ when using the Dark Energy Survey Year 5 (DESY5) SNe data, and 1.3σ with PantheonPlus. In all the cases we consider, the derived DE equation of state parameter presents evidence for phantom crossing. By investigating potential systematic effects in the low-redshift samples of DESY5 observations, we confirm that correcting for the offset in apparent magnitude with respect to PantheonPlus data, as suggested in previous studies, completely removes the tension. Furthermore, we assess the risk of potentially overfitting the data by changing the number of interpolation nodes. As expected, we find that with lesser nodes we get a smoother reconstructed behavior of the DE density, although with similar overall features. The pipeline developed in this work is ready to be used with future high-precision data to further investigate the evidence for a non-standard background evolution. Copyright © 2025, The Authors. All rights reserved.

关键词： Supernovae

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：