检索结果-内蒙古大学图书馆

31st IEEE international conference on Computer Design (ICCD)

作者： van den Braak, Gert-Jan Gomez-Luna, Juan Corporaal, Henk Gonzalez-Linares, Jose Maria Guil, Nicolas Eindhoven Univ Technol Dept Elect Engn NL-5600 MB Eindhoven Netherlands Univ Cordoba Dept Comp Architecture & Elect E-14071 Cordoba Spain Univ Malaga Dept Comp Architecture E-29071 Malaga Spain

ISBN: (纸本)9781479929870

GPUs are increasingly used as compute accelerators. With a large number of cores executing an even larger number of threads, significant speed-ups can be attained for parallel workloads. applications that rely on atomic operations, such as histogram and Hough transform, suffer from serialization of threads in case they update the same memory location. Previous work shows that reducing this serialization with software techniques can increase performance by an order of magnitude. We observe, however, that some serialization remains and still slows down these applications. Therefore, this paper proposes to use a hash function in both the addressing of the banks and the locks of the scratchpad memory. To measure the effects of these changes, we first implement a detailed model of atomic operations on scratchpad memory in GPGPU-Sim, and verify its correctness. Second, we test our proposed hardware changes. They result in a speed-up up to 4.9x and 1.8x on implementations utilizing the aforementioned software techniques for histogram and Hough transform applications respectively, with minimum hardware costs.

关键词： graphics processing units multi-threading parallel architectures random-access storage

来源：评论

学校读者我要写书评

暂无评论

Collecting Clustering Coefficient of distributed Graph Data with Shuffled Differential Privacy 22

Collecting Clustering Coefficient of Distributed Graph Data ...

引用

22nd IEEE international Symposium on parallel and distributed processing with applications, ISPA 2024

作者： Ding, Hongfa Fu, Peiwang Luo, Yingxuan Jiang, Heling Liu, Hai Guizhou University of Fiance and Economics School of Information Guiyang China Co. Ltd Guiyang China Jiangxi Datang International Fuzhou Power Generation Co. Ltd Fuzhou China

ISBN: (纸本)9798331509712

The intricate properties and relevance of graph data make it difficult to collect graph statistics privately via differential privacy (DP). Traditional centralized or local DP on graph data, face challenges like third-party threats and low data utility when collecting the clustering coefficient. In this regard, we introduce GCC-SDP, a scheme for collecting distributed Graph Clustering Coefficient with Shuffled DP (SDP). GCC-SDP gathers the local wedge lists of all edges and adjacency bit vectors through SDP and random response for calculating the noisy local triangle counts. It then collects the local degree values of all users by using Laplace mechanism, followed by estimating the global clustering coefficient of the global graph data by data collector. We provide specific steps of GCC-SDP and demonstrate through theoretical analysis that GCC-SDP conforms to various DPs, with unbiased results. Empirical experiments show that GCC-SDP performs better than existing local DP-based techniques across most accuracy metrics. © 2024 IEEE.

关键词： Data collection

来源：评论

学校读者我要写书评

暂无评论

Classification and. Advantages parallel Computing in Process Computation : A Systematic Literature Review 4

Classification and. Advantages Parallel Computing in Process...

引用

4th international conference on Computing, Engineering, and Design (ICCED)

作者： Fernando, Erick Murad, Dina Fitria Wijanarko, Bambang Dwi Bina Nusantara Univ Sch Informat Syst Informat Syst Dept Jakarta 11480 Indonesia Bina Nusantara Univ Binus Online Learning Informat Syst Dept Jakarta Indonesia Bina Nusantara Univ Binus Online Learning Comp Sci Dept Jakarta 11480 Indonesia

ISBN: (纸本)9781538693780

Data Management requires computing devices DIM can perform data processes to form better information. With the development of data, the processor can be done with one unit only, over time required computing devices that have high performance. parallel Computing is one of the techniques of doing computing simultaneously by utilizing several independent computers simultaneously. parallel computers can be grouped according to the level at which hardware supports parallelism This classification is generally imalogous to the distance between basic computing nodes. This research will focus on looking at the widely used classification trends in this parallelism that affect the performance of these calculations. This study uses a systematic literature review to find many classifications in parallel computing. literature is taken from a reputable journal database is ACM Digital Library, IEEE Xplore Digital Library, Science Diroct, Ernorald Insight. The results of this study are mostly conducted in the United States and China so as to provide many contributions. classification of parallelism, mostly done in parallel computing include distributed parallel, Multi-Core Processor, Massively parallel Computing, and Graph processing Unit (GPU). In this study also illustrates the advantages in the application of computer parallel based on its classification. In essence the advantages in the application of computer parallel improve performance performance, as well as effective and efficiency in a process that is done

关键词： parallel Computation parallel Computer Process Computation SLR

来源：评论

学校读者我要写书评

暂无评论

Does IDF improve declustering techniques for keyword-based information retrieval?

Does IDF improve declustering techniques for keyword-based i...

引用

20th IASTED international conference on parallel and distributed Computing and Systems, PDCS 2008

作者： Behl, Sanjiv Verma, Rakesh M. Computer Science Department University of Houston-Victoria Victoria TX 77901 United States Computer Science Department University of Houston Houston TX 77204 United States

ISBN: (纸本)9780889867741

Multiple-disk architectures are an attractive approach to meet high performance I/O demands in I/O intensive applications such as search engines, web servers and information retrieval systems. This requires that the issues of dynamic load balancing and access parallelism be addressed, which is the goal of this paper. In this paper we address the problem of document declus-tering in a keyword-based information retrieval system for parallel architectures consisting of a single processor and multiple disks. In the vector-space information retrieval model the inverse-document-frequency factor was found to improve the query performance. We study whether this is the case for declustering as well. We propose and evaluate experimentally four similarity-based methods, viz., vector, euclidean, vector without idf, and euclidean without idf for declus-tering documents. Interestingly, our results show that the vector method significantly outperforms the vector without idf, but the euclidean method is only slightly superior to the euclidean method without idf in one scenario and too close to call in another scenario. The vector method is the best for the so-called simple plan. We also introduce a highest-frequency first retrieval scenario and compare the methods under this scenario, and find the methods are too close to call in this case.

关键词： Vectors

来源：评论

学校读者我要写书评

暂无评论

Config 2.0: Towards reinforcement learning based configuration of stream processing systems 21

Config 2.0: Towards reinforcement learning based configurati...

引用

22nd international Middleware conference, Middleware 2021

作者： Tahir, Jawad Technical University of Munich Garching Germany

ISBN: (纸本)9781450391559

distributed stream processing systems (DSPSs) have enabled us to build scalable, fast, and reactive applications. As faults are common in any distributed system, DSPSs allow for recovering from faults using various techniques, for example, snapshotting their state on a key-value (KV) store and then replaying the events in case of a failure. The performance of the recovery process is dependent on the workload, configuration of the DSPS and performance of the KV store. Performance tuning of a KV store is a notoriously difficult job. In this thesis proposal, we propose a reinforcement learning (RL) based agent which configures the DSPS and the KV store under varying workloads to minimize the impact of a failure on the performance. © 2021 ACM.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Microstructural and mechanical characterization of parallel layered WC- NiCr weld overlay on 080 M40 steel substrate prepared using additive manufacturing

Microstructural and mechanical characterization of parallel ...

引用

5th international conference on Advances in Steel, Power and Construction Technology

作者： Kumar, K. Varun Phani, M. Kalyan OP Jindal Univ Dept Met Engn Raigarh 496109 Chhattisgarh India

080 M40 Medium carbon steel has potential applications in axles, shafts, bolts, studs, spindles, automotive components and many more. In high friction applications these steels possess better wear resistance, and after weld overlay, they can exhibit superior properties. Additive manufacturing (AM) is an advanced method in which material in the form of powder or wire is used to form a desired shape. Laser Metal Deposition (LMD) is an AM method which is used in industrial application to recreate, repair and produce corrosion resistance weld overlay on a part. In comparison with conventional overlay techniques such as Plasma Transfer Arc (PTA), Metal Inert Gas (MIG) & Tungsten Inert Gas (TIG);Laser weld overlays are faster techniques which can make thin sections with good metallurgical bond and with high productivity and flexibility. Apart from that it has also advantage of low heat affected zone, minimum metal dilution, and can produce thin layers with good aesthetics. In this paper, 080 M40 steel samples were used as substrate and WC- NiCr powder was weld overlayed using LMD method. Such kind of weld overlays will be had potential application in the field of mining. The weld overlay was deposited in parallel layered fashion and effect of set of parameters like feed rate, dilution, size of the bead and processing speed were used to produce weld overlays of 1 mm thickness. metallurgical, micro structural and mechanical properties of these weld overlay materials have been investigated. These properties are very important in the processing of mining tools as well as other parts produced using the LMD process and also for other future applications. Further effect of laser processing parameters on the resulting properties of these steels weld overlays are discussed. (c) 2022 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the 5th international conference on Advances in Steel, Power and Construction Technology

关键词： Laser metal deposition Additive manufacturing Laser weld overlay Carbide distribution WC-NiCr

来源：评论

学校读者我要写书评

暂无评论

Automatic Low-Overhead Load-Imbalance Detection in MPI applications 27th

Automatic Low-Overhead Load-Imbalance Detection in MPI Appli...

引用

27th international European conference on parallel and distributed Computing (Euro-Par)

作者： Arzt, Peter Fischler, Yannic Lehr, Jan-Patrick Bischof, Christian Tech Univ Darmstadt Dept Comp Sci Sci Comp Darmstadt Germany

ISBN: (纸本)9783030856656;9783030856649

Load imbalances are a major reason for efficiency loss in highly parallel applications. Hence, their identification is of high relevance in performance analysis and tuning. We present a low-overhead approach to automatically identify load-imbalanced regions and filter out irrelevant ones based on new selection heuristics in our PIRA tool for automatic instrumentation refinement for the Score-P measurement system. For the LULESH mini-app as well as the Ice-sheet and Sea-level System Model simulation package we, thus, correctly identify existing load imbalances while maintaining a runtime overhead of less than 10% for all but one input. Moreover, the traces generated are suitable for Scalasca's automatic trace analysis.

关键词： Score-P Load imbalance HPC Performance analysis

来源：评论

学校读者我要写书评

暂无评论

Task-Based Programming on Emerging parallel Architectures for Finite-Differences Seismic Numerical Kernel 24th

Task-Based Programming on Emerging Parallel Architectures fo...

引用

24th international European conference on parallel and distributed Computing (Euro-Par)

作者： Moustafa, Salli Kirschenmann, Wilfried Dupros, Fabrice Aochi, Hideo ANEO Boulogne France Bur Rech Geol & Minieres Orleans France

ISBN: (纸本)9783319969831;9783319969824

In recent years, heterogeneous hardware have generalized in almost all supercomputer nodes, requiring a profound shift on the way numerical applications are implemented. This paper, illustrates the design and implementation of a seismic wave propagation simulator, based on the finite-differences numerical scheme, and specifically tailored for such massively parallel hardware infrastructures. The application data-flow is built on top of PaRSEC, a generic task-based runtime system. The numerical kernels, designed for maximizing data reuse can efficiently leverage large SIMD units available in modern CPU cores. A strong scalability study on a cluster of Intel KNL processors illustrates the application performances.

关键词： High-performance computing C plus plus generic programming SIMD Task-based runtime system PaRSEC Seismic wave propagation

来源：评论

学校读者我要写书评

暂无评论

An XML gateway to patient data for medical research applications

引用

techniques IN BIOINFORMATICS AND MEDICAL INFORMATICS 2002年第1期980卷 236-246页

作者： Bui, AAT Weinger, GS Barretta, SJ Dionisio, JDN Kangarloo, H Univ Calif Los Angeles Telemed Div Dept Radiol Los Angeles CA 90024 USA

As the medical environment becomes increasingly electronic, clinical databases are continually growing, accruing masses of patient information. This wealth of data is an invaluable source of information to researchers, serving as a testbed for the development of new information technologies and as a repository of real-world data for data mining and population-based studies. However, the true utility of this information is not fulfilled, in part because of issues pertaining to security and patient confidentiality, but also due to the lack of an effective infrastructure to access the data. This paper describes a system, DataServer, that permits researchers to query and retrieve data from multiple clinical data sources, automatically deidentifying patient data so that it can be used for research purposes. DataServer functions as an application framework, enabling extensible markup language (XML)-based querying of existing medical databases. Key aspects of DataServer include ready inclusion of new information resources, minimal processing impact on existing clinical systems via a distributed cache, and flexible output representation via XSL (eXtensible Style Language) transforms.

关键词： XML distributed databases patient deidentification

来源：评论

学校读者我要写书评

暂无评论

Universal sigma-delta ADC for intelligent distributed instrumentation 4

Universal sigma-delta ADC for intelligent distributed instru...

引用

4th IMEKO TC4 conference on Advanced A/D and D/A Conversion techniques and Their applications and the 7th Workshop on ADC Modelling and Testing 2002

作者： Kochan, R. Sachenko, A. Kochan, V. Vasylkiv, N. Institute of Computer Information Technologies Ternopil Academy of National Economy Ukraine

ISBN: (纸本)9781634391832

In this papers the precision single-board, 8 channels, 24 bit sigma-delta ADC with on-line remote reprogramming mode is described. The results of experimental researches of the noise immunity of the developed ADC in real (industrial) measurement conditions are presented. The results of experimental researches of the prototype in the same measurement conditions are also considered. Copyright © (2002) by the international Measurement Confederation (IMEKO).

关键词： Signal processing

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：