检索结果-内蒙古大学图书馆

International Conference on Parallel Processing (ICPP)

作者： Sameh Shohdy Abhinav Vishnu Gagan Agrawal Department of Computer Science and Engineering The Ohio State University Columbus OH High Performance Computing Group Pacific Northwest National Laboratory Richland WA

ISBN: (纸本)9781509028245

Support Vector Machines (SVM) is a popular Machine Learning algorithm, which is used for building classifiers and models. Parallel implementations of SVM, which can run on large scale supercomputers, are becoming commonplace. However, these supercomputers -- designed under constraints of data movement -- frequently observe faults in compute devices. Many device faults manifest as permanent process/node failures. In this paper, we present several approaches for designing fault tolerant SVM algorithms. First, we present an in-depth analysis to identify the critical data structures, and build baseline algorithms that simply periodically checkpoint these data structures. Next, we propose a novel algorithm, which requires no inter-node data movement for checkpointing, and only O(n2/p2) recovery time -- a small fraction of the expected O(n3/p) time-complexity of SVM. We implement these algorithms and evaluate them on a large scale cluster. Our evaluation indicates that the overall data movement for checkpointing in the baseline algorithm can be up to 100x the dataset size!, while the proposed novel algorithm is completely communication-free of checkpointing. In addition, it saves up to 20x space, while providing better (by an average of 5.5x speedup on 256 cores) recovery time than the baseline algorithm with different number of checkpoints. The experiments also show that our communication avoiding algorithm outperforms Spark MLLib SVM implementation by an average of 6.4x with 256 cores in the case of failure.

关键词： Support vector machines Algorithm design and analysis Checkpointing Fault tolerance Fault tolerant systems Data structures Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Fault Tolerant Frequent Pattern Mining

Fault Tolerant Frequent Pattern Mining

引用

International Conference on high performance computing

ISBN: (纸本)9781509054121

FP-Growth algorithm is a Frequent Pattern Mining (FPM) algorithm that has been extensively used to study correlations and patterns in large scale datasets. While several researchers have designed distributed memory FP-Growth algorithms, it is pivotal to consider fault tolerant FP-Growth, which can address the increasing fault rates in large scale systems. In this work, we propose a novel parallel, algorithm-level fault-tolerant FP-Growth algorithm. We leverage algorithmic properties and MPI advanced features to guarantee an O(1) space complexity, achieved by using the dataset memory space itself for checkpointing. We also propose a recovery algorithm that can use in-memory and disk-based checkpointing, though in many cases the recovery can be completed without any disk access, and incurring no memory overhead for checkpointing. We evaluate our FT algorithm on a large scale InfiniBand cluster with several large datasets using up to 2K cores. Our evaluation demonstrates excellent efficiency for checkpointing and recovery in comparison to the disk-based approach. We have also observed 20x average speed-up in comparison to Spark, establishing that a well designed algorithm can easily outperform a solution based on a general fault-tolerant programming model.

关键词： Fault tolerance Fault tolerant systems Checkpointing Algorithm design and analysis Programming Sparks Data structures

来源：评论

学校读者我要写书评

暂无评论

Unequal Failure Protection Coding Technology for Cloud Storage Systems

Unequal Failure Protection Coding Technology for Cloud Stora...

引用

IEEE International Conference on Cluster computing

作者： Yupeng Hu Yonghe Liu Wenjia Li Nong Xiao Zheng Qin Shu Yin College of Computer Science and Electronic Engineering Hunan University China Department of Computer Science and Engineering UT-Arlington USA Department of Computer Science New York Institute of Technology USA State Key Laboratory of High Performance Computing National University of Defense Technology China

ISBN: (纸本)9781509036547

In recent years, erasure codes have become the de facto standard for data protection of large scale distributed cloud storage systems at the cost of an affordable storage overhead. While traditional erasure coding schemes, such as Reed-Solomon codes, suffer from high reconstruction cost and I/Os. The recent past has seen a plethora of efforts to optimize the tradeoff between the reconstruction cost, I/Os and storage overhead. Quietly different from all prior studies, in this paper, our erasure coding technology makes the first attempt to take advantage of the unequal failure rates across the disks/nodes to optimize the reconstruction performance and system reliability. Specifically, our proposed technology, the Unequal Failure Protection based Local Reconstruction Code (UFP-LRC) divides the data blocks into several unequal-sized groups with local parities, assigning the data blocks stored on more failure-prone disks/nodes into the smaller-sized group, so as to provide unequal failure protection for each group. In this way, by exploiting the nonuniform local parity degrees, the proposed UFP-LRC enables the data blocks that are stored on more failure-prone disks/nodes to tolerate a greater number of failures while suffer from less repair cost than others, leading to a substantial improvement of overall repair performance and reliability for cloud storage system. We perform numerical analysis and build a prototype storage system to verify our approach. The analytical results show that the UFPLRC technique gradually outperforms LRC along the increase of failure rate ratio. Also, extensive experiments show that, when compared to LRC, UFP-LRC is able to achieve a 10% to 13% improvement in throughput, and a 8% to 12% reduction in decoding latency, while retaining a comparable overall reliability.

关键词： Cloud computing Encoding Maintenance engineering computer science Facebook Redundancy

来源：评论

学校读者我要写书评

暂无评论

Quantum key distribution with distinguishable decoy states

arXiv

引用

arXiv 2017年

作者： Huang, Anqi Sun, Shi-Hai Liu, Zhihong Makarov, Vadim Institute for Quantum Information State Key Laboratory of High Performance Computing College of Computer National University of Defense Technology Changsha410073 China Institute for Quantum Computing University of Waterloo WaterlooONN2L 3G1 Canada Department of Electrical and Computer Engineering University of Waterloo WaterlooONN2L 3G1 Canada College of Science National University of Defense Technology Changsha410073 China College of Mechatronic Engineering and Automation National University of Defense Technology Changsha410073 China Russian Quantum Center MISIS University Moscow Department of Physics and Astronomy University of Waterloo WaterlooONN2L 3G1 Canada

The decoy state protocol has been considered to be one of the most important methods to protect the security of quantum key distribution (QKD) with a weak coherent source. Here we test two experimental approaches to generating the decoy states with different intensities: modulation of the pump current of a semiconductor laser diode, and external modulation by an optical intensity modulator. The former approach shows a side-channel in the time domain that allows an attacker to distinguish s signal state from a decoy state, breaking a basic assumption in the protocol. We model a photon-number-splitting attack based on our experimental data, and show that it compromises the system's security. Then, based on the work of K. Tamaki et al. [New J. Phys. 18, 065008 (2016)], we obtain two analytical formulas to estimate the yield and error rate of single-photon pulses when the signal and decoy states are distinguishable. The distinguishability reduces the secure key rate below that of a perfect decoy-state protocol. To mitigate this reduction, we propose to calibrate the transmittance of the receiver (Bob's) unit. We apply our method to three QKD systems and estimate their secure key rates. Copyright © 2017, The Authors. All rights reserved.

关键词： Modulation

来源：评论

学校读者我要写书评

暂无评论

Parallel Erasure Coding: Exploring Task Parallelism in Erasure Coding for Enhanced Bandwidth and Energy Efficiency

Parallel Erasure Coding: Exploring Task Parallelism in Erasu...

引用

International Conference on Networking, Architecture, and Storage (NAS)

作者： Hsing-bung Chen Song Fu High Performance Computing-Design Group Los Alamos National Laboratory Los Alamos New Mexico USA Department of Computer Science and Engineering University of North Texas Denton Texas USA

ISBN: (纸本)9781509033164

Very large data sets within the range of megabytes to terabytes generated daily from checkpoint-and- restart processes are seen in today's scientific simulations. Reliability and durability are two important factors to build an archive storage system. Erasure code based object storage systems are becoming popular choices for archive storage systems due to cost-effective storage space saving schemes and higher fault-resilience capabilities. Both erasure code encoding and decoding procedures involve heavy array, matrix, and table-lookup compute intensive operations. Current solutions of the erasure coding process are based on single process approach which is not capable of processing very large data sets efficient and effectively. In this paper, we address the bottleneck problem of single process erasure encoding by leveraging task parallelism offered by multi-core computers. We add parallel processing capability to the erasure coding process. More specifically, we develop a parallel erasure coding software, called parEC. It explores the MPI run time parallel I/O environment and integrates data placement process for distributing encoded data blocks to destination storage devices. We evaluate the performance of parEC in terms of both encoding throughput and energy efficiency. We also compare the performance of two task scheduling algorithms for parEC. Our experimental results show parEC can significantly reduce the encoding time (i.e., by 74.06%-96.86%) and energy consumption (i.e., by 73.57%-96.86%), and Demand-based Workload Assignment (DBWA) algorithm can a high system utilization (i.e., 95.23%).

关键词： Encoding Multicore processing Parallel processing Servers Testing Software systems

来源：评论

学校读者我要写书评

暂无评论

A novel PUF based on cell error rate distribution of STT-RAM

A novel PUF based on cell error rate distribution of STT-RAM

引用

Asia and South Pacific Design Automation Conference

作者： Xian Zhang Guangyu Sun Yaojun Zhang Yiran Chen Hai Li Wujie Wen Jia Di Center for Energy-Efficient Computing and Applications Peking University Beijing China Collaborative Innovation Center of High Performance Computing NUDT Changsha China Department of Electrical and Computer Engineering University of Pittsburgh Pittsburgh PA USA Department of Electrical and Computer Engineering Florida International University Miami FL USA Computer Science and Computer Engineering Department University of Arkansas Fayetteville AR USA

ISBN: (纸本)9781467395700

Physical Unclonable Functions (PUFs) have been widely proposed as security primitives to provide device identification and authentication. Recently, PUFs based on Non-volatile Memory (NVM) are widely proposed since the promise of NVMs' wide application. In addition, NVM-based PUFs are considered to be more immune to invasive attack and simulation attack than CMOS-based PUFs. However, the existing NVM-based PUF either shows the unreliability under environmental variations or need extra modifications to the IC manufacturing process. In this work, we propose err-PUF, a novel PUF design based on the cell error rate distribution of STT-RAM. Instead of using the distribution directly, we generate a stable fingerprint based on a novel concept called Error-rate Differential Pair (EDP) without modifications to the read/write circuits. Comprehensive results demonstrate that err-PUF can achieve sufficient reliability under environmental variations, which can significantly impact the cell error rates. Moreover, compared with existing approaches, err-PUF has a higher speed and lower power consumption with negligible overhead.

关键词： Nonvolatile memory Bit error rate Robustness

来源：评论

学校读者我要写书评

暂无评论

Erratum to: Search for supersymmetry in final states with two same-sign or three leptons and jets using 36 fb−1 of = 13 TeV pp collision data with the ATLAS detector

引用

Journal of high Energy Physics 2019年第8期2019卷 1-19页

作者： Aaboud, M. Aad, G. Abbott, B. Abdinov, O. Abeloos, B. Abidi, S. H. AbouZeid, O. S. Abraham, N. L. Abramowicz, H. Abreu, H. Abreu, R. Abulaiti, Y. Acharya, B. S. Adachi, S. Adamczyk, L. Adelman, J. Adersberger, M. Adye, T. Affolder, A. A. Afik, Y. Agatonovic-Jovin, T. Agheorghiesei, C. Aguilar-Saavedra, J. A. Ahlen, S. P. Ahmadov, F. Aielli, G. Akatsuka, S. Akerstedt, H. Åkesson, T. P. A. Akilli, E. Akimov, A. V. Alberghi, G. L. Albert, J. Albicocco, P. Alconada Verzini, M. J. Alderweireldt, S. C. Aleksa, M. Aleksandrov, I. N. Alexa, C. Alexander, G. Alexopoulos, T. Alhroob, M. Ali, B. Aliev, M. Alimonti, G. Alison, J. Alkire, S. P. Allbrooke, B. M. M. Allen, B. W. Allport, P. P. Aloisio, A. Alonso, A. Alonso, F. Alpigiani, C. Alshehri, A. A. Alstaty, M. I. Alvarez Gonzalez, B. Álvarez Piqueras, D. Alviggi, M. G. Amadio, B. T. Amaral Coutinho, Y. Amelung, C. Amidei, D. Amor Dos Santos, S. P. Amoroso, S. Amundsen, G. Anastopoulos, C. Ancu, L. S. Andari, N. Andeen, T. Anders, C. F. Anders, J. K. Anderson, K. J. Andreazza, A. Andrei, V. Angelidakis, S. Angelozzi, I. Angerami, A. Anisenkov, A. V. Anjos, N. Annovi, A. Antel, C. Antonelli, M. Antonov, A. Antrim, D. J. Anulli, F. Aoki, M. Aperio Bella, L. Arabidze, G. Arai, Y. Araque, J. P. Araujo Ferraz, V. Arce, A. T. H. Ardell, R. E. Arduh, F. A. Arguin, J-F. Argyropoulos, S. Arik, M. Armbruster, A. J. Armitage, L. J. Arnaez, O. Arnold, H. Arratia, M. Arslan, O. Artamonov, A. Artoni, G. Artz, S. Asai, S. Asbah, N. Ashkenazi, A. Asquith, L. Assamagan, K. Astalos, R. Atkinson, M. Atlay, N. B. Augsten, K. Avolio, G. Axen, B. Ayoub, M. K. Azuelos, G. Baas, A. E. Baca, M. J. Bachacou, H. Bachas, K. Backes, M. Bagnaia, P. Bahmani, M. Bahrasemani, H. Baines, J. T. Bajic, M. Baker, O. K. Baldin, E. M. Balek, P. Balli, F. Balunas, W. K. Banas, E. Bandyopadhyay, A. Banerjee, Sw. Bannoura, A. A. E. Barak, L. Barberio, E. L. Barberis, D. Barbero, M. Barillari, T. Barisits, M-S Barkeloo, J. T. Barklow, T. Barlow, N. Barnes, S. L. Barnett, B. M. Barnett, R. M. Barnovska-Blenessy, Z Faculté des Sciences Université Mohamed Premier and LPTPM Oujda Morocco CPPM Aix-Marseille Université and CNRS/IN2P3 Marseille France Homer L. Dodge Department of Physics and Astronomy University of Oklahoma Norman U.S.A. Institute of Physics Azerbaijan Academy of Sciences Baku Azerbaijan LAL Univ. Paris-Sud CNRS/IN2P3 Université Paris-Saclay Orsay France Department of Physics University of Toronto Toronto Canada Santa Cruz Institute for Particle Physics University of California Santa Cruz Santa Cruz U.S.A. Department of Physics and Astronomy University of Sussex Brighton United Kingdom Raymond and Beverly Sackler School of Physics and Astronomy Tel Aviv University Tel Aviv Israel Department of Physics Technion: Israel Institute of Technology Haifa Israel Center for High Energy Physics University of Oregon Eugene U.S.A. Department of Physics Stockholm University Stockholm Sweden The Oskar Klein Centre Stockholm Sweden INFN Gruppo Collegato di Udine Udine Italy ICTP Trieste Italy Department of Physics King’s College London London United Kingdom International Center for Elementary Particle Physics and Department of Physics The University of Tokyo Tokyo Japan AGH University of Science and Technology Faculty of Physics and Applied Computer Science Krakow Poland Department of Physics Northern Illinois University DeKalb U.S.A. Fakultät für Physik Ludwig-Maximilians-Universität München München Germany Particle Physics Department Rutherford Appleton Laboratory Didcot United Kingdom Institute of Physics University of Belgrade Belgrade Serbia Department of Physics Alexandru Ioan Cuza University of Iasi Iasi Romania Laboratório de Instrumentação e Física Experimental de Partículas – LIP Lisboa Portugal Departamento de Fisica Teorica y del Cosmos Universidad de Granada Granada Portugal Department of Physics Boston University Boston U.S.A. Joint Institute for Nuclear Research JINR Dubna Dubna Russia INFN Sezione di Roma Tor Vergata Roma Italy Dip

One correction is made to the figure 4e of the paper.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An emergent space for distributed data with hidden internal order through manifold learning

arXiv

引用

arXiv 2017年

作者： Kemeth, Felix P. Haugland, Sindre W. Dietrich, Felix Bertalan, Tom Höhlein, Kevin Li, Qianxiao Bollt, Erik M. Talmon, Ronen Krischer, Katharina Kevrekidis, Ioannis G. Physik-Department Nonequilibrium Chemical Physics Technische Universität München James-Franck-Str. 1 GarchingD-85748 Germany Institute for Advanced Study Technische Universität München Lichtenbergstr. 2a GarchingD-85748 Germany Department of Chemical and Biological Engineering Princeton University PrincetonNJ08544 United States Department of Chemical and Biomolecular Engineering Department of Applied Mathematics and Statistics Johns Hopkins University JHMI Department of Mechanical Engineering MIT CambridgeMA02139 United States Institute of High Performance Computing 1 Fusionopolis Way #16-16 Connexis North Singapore138632 Singapore Department of Mathematics Department of Electrical and Computer Engineering Clarkson Center for Complex Systems Science Clarkson University PotsdamNY13699-5815 United States Department of Electrical Engineering Technion Israel Institute of Technology Technion City Haifa32000 Israel

Manifold-learning techniques are routinely used in mining complex spatiotemporal data to extract useful, parsimonious data representations/parametrizations;these are, in turn, useful in nonlinear model identification tasks. We focus here on the case of time series data that can ultimately be modelled as a spatially distributed system (e.g. a partial differential equation, PDE), but where we do not know the space in which this PDE should be formulated. Hence, even the spatial coordinates for the distributed system themselves need to be identified - to "emerge from"- the data mining process. We will first validate this "emergent space" reconstruction for time series sampled without space labels in known PDEs;this brings up the issue of observability of physical space from temporal observation data, and the transition from spatially resolved to lumped (order-parameter-based) representations by tuning the scale of the data mining kernels. We will then present actual emergent space "discovery" illustrations. Our illustrative examples include chimera states (states of coexisting coherent and incoherent dynamics), and chaotic as well as quasiperiodic spatiotemporal dynamics, arising in partial differential equations and/or in heterogeneous networks. We also discuss how datadriven "spatial" coordinates can be extracted in ways invariant to the nature of the measuring instrument. Such gauge-invariant data mining can go beyond the fusion of heterogeneous observations of the same system, to the possible matching of apparently different systems. For an older version of this article, including other examples, see https://***/abs/1708.05406. Copyright © 2017, The Authors. All rights reserved.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Optimizing System Call Latency of ARM Virtual Machines

引用

Journal of Physics: Conference Series 2017年第1期787卷

作者： Song-Woo Sok Young-Woo Jung Cheol-Hun Lee High-Performance Computing Dept Electronics and Telecommunications Research Institute 218 Gajeong-ro Yuseong-gu Daejeon 34129 Korea Department of Computer Science & Engineering Chungnam National University 99 Daehak-ro Yuseong-gu Daejeon 34134 Korea

This paper introduces ViMo-S, a type 1 hypervisor for ARMv7 and ARMv8-based ARM server systems. It supports full virtualization to run existing operating systems and applications unmodified. It uses ARM hardware virtualization extensions to optimize the performance of virtual machines, especially system call latency. Therefore, its virtual machines' system call latency is near physical machine's, while other hypervisors like Xen and KVM show relatively slower and unstable performances in benchmark tests.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Comment on 'Inherent security of phase coding quantum key distribution systems against detector blinding attacks' (2018 Laser Phys. Lett. 15 095203)

引用

Laser Physics Letters 2018年第1期16卷

作者： Aleksey Fedorov Ilja Gerhardt Anqi Huang Jonathan Jogenfors Yury Kurochkin Antía Lamas-Linares Jan-Åke Larsson Gerd Leuchs Lars Lydersen Vadim Makarov Johannes Skaar Russian Quantum Center Skolkovo Moscow 143025 Russia QRate Skolkovo Moscow 143025 Russia QApp Skolkovo Moscow 143025 Russia Institute of Physics University of Stuttgart and Institute for Quantum Science and Technology Pfaffenwaldring 57 D-70569 Stuttgart Germany Max Planck Institute for Solid State Research Heisenbergstraße 1 D-70569 Stuttgart Germany Institute for Quantum Information & State Key Laboratory of High Performance Computing College of Computer National University of Defense Technology Changsha 410073 People's Republic of China Department of Electrical Engineering Linköping University SE-58183 Linköping Sweden Texas Advanced Computing Center The University of Texas at Austin Austin Texas United States of America Max Planck Institute for the Science of Light and University of Erlangen-Nürnberg D-91058 Erlangen Germany Kringsjåvegen 3E NO-7032 Trondheim Norway National University of Science and Technology MISIS Moscow 119049 Russia Department of Technology Systems University of Oslo Box 70 NO-2027 Kjeller Norway

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：