检索结果-内蒙古大学图书馆

20th IFIP WG 10.3 International Conference on Network and parallel Computing, NPC 2024

作者： Guo, Mingfeng Deng, Liang Dai, Zhe Li, Ruitian Lin, Gaofeng Liu, Jie Computational Aerodynamics Institute China Aerodynamics Research and Development Center Mianyang China Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha China

ISBN: (纸本)9789819628292

Sparse triangular solve (SpTRSV) is a vital component in various scientific applications, and numerous GPU-based SpTRSV algorithms have been proposed. Synchronization-free SpTRSV is currently the mainstream algorithm on GPU due to its short preprocessing time and outstanding performance. However, we observed that this algorithm still has two performance bottlenecks. Firstly, the thread-level parallel mode can introduce to thread divergence issues within GPU warps during the writing phase. Secondly, the thread-level and warp-level fusion mode may struggles to fully exploit GPU resources due to suboptimal mapping relationships between rows and threads. To address these issues, this paper proposes DaCPSpTRSV, a new synchronization-free algorithm with GPU-friendly data communication and parallelism strategies. Specifically, we first develop a fast-forward thread-level approach, incorporating an efficient global memory access pattern and a light-weight dependency control mechanism, to optimize data communication and alleviate thread divergence. A fine-grained fusion strategy is then proposed to maximize GPU parallelism by adaptively selecting the suitable thread-level or warp-level modes. Moreover, the commonly-used compressed sparse row (CSR) format is employed in our DaCPSpTRSV, enhancing the versatility of our algorithm. We evaluate our approach using 245 matrices from the SuiteSparse Matrix Collection on two NVIDIA GPUs, demonstrating speedup ratios of up to 4.77×, 4.94×, 1.67×, and 1.62× compared to cuSPARSE, Sync-Free, CapelliniSpTRSV, and YuenyeungSpTRSV, respectively. The project is open-sourced at https://***/gmfff12334/DaCP. © IFIP International Federation for Information processing 2025.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

A data representation method using distance correlation

引用

Frontiers of Computer Science 2025年第1期19卷 1-14页

作者： Xinyan LIANG Yuhua QIAN Qian GUO Keyin ZHENG Institute of Big Data Science and Industry Shanxi UniversityTaiyuan 030006China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education Shanxi UniversityTaiyuan 030006China School of Computer Science and Technology Taiyuan University of Science and TechnologyTaiyuan 030024China Shanxi Key Laboratory of Big Data Analysis and Parallel Computing Taiyuan University of Science and TechnologyTaiyuan 030024China

Association in-between features has been demonstrated to improve the representation ability of data. However, the original association data reconstruction method may face two issues: the dimension of reconstructed data is undoubtedly higher than that of original data, and adopted association measure method does not well balance effectiveness and efficiency. To address above two issues, this paper proposes a novel association-based representation improvement method, named as AssoRep. AssoRep first obtains the association between features via distance correlation method that has some advantages than Pearson’s correlation coefficient. Then an improved matrix is formed via stacking the association value of any two features. Next, an improved feature representation is obtained by aggregating the original feature with the enhancement matrix. Finally, the improved feature representation is mapped to a low-dimensional space via principal component analysis. The effectiveness of AssoRep is validated on 120 datasets and the fruits further prefect our previous work on the association data reconstruction.

关键词： association representation distance correlation classification

来源：评论

学校读者我要写书评

暂无评论

TFS: Revisiting Temporal Language Grounding from Frequency Spiking Perspective

TFS: Revisiting Temporal Language Grounding from Frequency S...

引用

International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Yifan Lyu Hongzhou Wu Lixiang Liu Chuxiong Sun Institute of Software Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China Parallel and Distributed Processing Lab National University of Defense Technology Changsha China Institute of Software Chinese Academy of Sciences Beijing China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Temporal Language Grounding (TLG) aims to localize moments in untrimmed videos that are most relevant to natural language queries. While existing weakly-supervised methods have achieved significant success in exploring cross-modal relationships, they still face a critical bottleneck: the interference of task-irrelevant information in query embeddings. To address this issue, we propose TLG Frequency Spiking (TFS), a dimensional mask derived from the frequency domain that models the varying importance specific to different queries. By enhancing the understanding of queries, TFS effectively optimizes the cross-modal alignment of visual and textual modalities. Experimental results show that TFS significantly outperforms state-of-the-art baselines on both the Charades-STA and ActivityNet-Captions datasets.

关键词： Visualization Grounding Frequency-domain analysis Neural networks Natural languages Interference Signal processing Speech processing Faces Videos

来源：评论

学校读者我要写书评

暂无评论

Exploiting Loop parallelism with Redundant Execution

引用

Journal of Computer Science & Technology 1997年第2期12卷 105-112页

作者：唐卫宇施武臧斌宇朱传琪 Institute of Parallel Processing Fudan University Shanghai

In this paper, a new loop transformation is proposed that can exploit parallelism in loops which cannot be found by traditional methods. Then the method is extended to show how to achieve maximum speedup of loops if there are infinite processors and how to balance the workload of parallel sections in loops i f there is fixed number of processors.

关键词： parallel compiler loop transformation parallel section maximum speedup

来源：评论

学校读者我要写书评

暂无评论

A New Approach to Pointer Analysis for Assignments

引用

Journal of Computer Science & Technology 2001年第3期16卷 242-250页

作者：黄波藏斌宇李靖朱传琪 Institute of Parallel Processing Fudan University Shanghai P.R. China

Pointer analysis is a technique to identify at compile-time the potential values of the pointer expressions in a program, which promises significant benefits for optimizing and parallelizing compilers. In this paper, a new approach to pointer analysis for assignments is presented. In this approach, assignments are classified into three categories: pointer assignments, structure (union) assignments and normal assignments which don't affect the point-to information. Pointer analyses for these three kinds of assignments respectively make up the integrated algorithm. When analyzing a pointer assignment, a new method called expression expansion is used to calculate both the left targets and the right targets. The integration of recursive data structure analysis into pointer analysis is a significant originality of this paper, which uniforms the pointer analysis for heap variables and the pointer analysis for stack variables. This algorithm is implemented in Agassiz, an analyzing tool for C programs developed by institute of parallel processing, Fudan University. Its accuracy and effectiveness are illustrated by experimental data.

关键词： pointer analysis pointer assignment point-to information recursive data structure

来源：评论

学校读者我要写书评

暂无评论

Service fault tolerance for highly reliable service-oriented systems: an overview

引用

Science China(Information Sciences) 2015年第5期58卷 7-18页

作者： ZHENG ZiBin LYU Michael Rung Tsong WANG HuaiMin Shenzhen Research Institute The Chinese University of Hong Kong National Laboratory for Parallel & Distributed Processing National University of Defense Technology

Service-oriented systems are widely-employed in e-business, e-government, finance, management systems, and so on. Service fault tolerance is one of the most important techniques for building highly reliable service-oriented systems. In this paper, we provide an overview of various service fault tolerance techniques,including sections on fault tolerance strategy design, fault tolerance strategy selection, and Byzantine fault tolerance. In the first section, we introduce the design of static and dynamic fault tolerance strategies, as well as the major problems when designing fault tolerance strategies. After that, based on various fault tolerance strategies, in the second section, we identify significant components from a complex service-oriented system, and investigate algorithms for optimal fault tolerance strategy selection. Finally, in the third section, we discuss a special type of service fault tolerance techniques, i.e., the Byzantine fault tolerance.

关键词： fault tolerance software reliability Web service SOA

来源：评论

学校读者我要写书评

暂无评论

SELF-ROUTING CONTROL ALGORITHMS AND THE PASSABILITY OF RANDOM INPUTS BY THE BASE-LINE NETWORK

引用

INFORMATION SCIENCES 1987年第1-2期43卷 139-154页

作者： ALHALLAQ, AY LAKSHMIVARAHAN, S Parallel Processing Institute School of Electrical Engineering and Computer Science University of Oklahoma Norman Oklahoma 7301 USA

The average bandwidth of (complete) cross-bar and multistage interconnection networks built out of smaller cross-bars under random inputs have been analyzed by a number of authors. This approach, while very useful, has a basic difficulty. Once the input to the network is generated (randomly or otherwise), since the network control algorithms are deterministic, this average-case approach does not shed light on what happens to any specific random input, but merely characterizes the ensemble average. In this paper, we first define new passability condition which includes the now well-known condition of passability of permutations as a special case. Using this condition, we then analyze the passability properties of different subclasses of random inputs. It is shown that the base-line network (which is a representative of an interesting class of unique path networks) is very robust in the sense that it can pass a variety of random inputs in much the same way as it handles the set of passable permutations.

关键词： COMPUTER SYSTEMS, DIGITAL

来源：评论

学校读者我要写书评

暂无评论

Efficient dictionary-based text rewriting using subsequential transducers

引用

Natural Language Engineering 2007年第4期13卷 353-381页

作者： Mihov, Stoyan Schulz, K.U. Institute for Parallel Processing Bulgarian Academy of Sciences Bulgaria CIS University of Munich Munich Germany

Problems in the area of text and document processing can often be described as text rewriting tasks: given an input text, produce a new text by applying some fixed set of rewriting rules. In its simplest form, a rewriting rule is given by a pair of strings, representing a source string (the "original") and its substitute. By a rewriting dictionary, we mean a finite list of such pairs;dictionary-based text rewriting means to replace in an input text occurrences of originals by their substitutes. We present an efficient method for constructing, given a rewriting dictionary D, a subsequential transducer ℑ that accepts any text t as input and outputs the intended rewriting result under the so-called "leftmost-longest match" replacement with skips, t′. The time needed to compute the transducer is linear in the size of the input dictionary. Given the transducer, any text t of length t is rewritten in a deterministic manner in time O( t + t′ ), where t′ denotes the resulting output text. Hence the resulting rewriting mechanism is very efficient. As a second advantage, using standard tools, the transducer can be directly composed with other transducers to efficiently solve more complex rewriting tasks in a single processing step. © 2007 Cambridge University Press.

关键词： Text processing

来源：评论

学校读者我要写书评

暂无评论

Evaluating the performance and scalability of MapReduce applications on X10 1

引用

Lecture Notes in Computer Science

作者： Zhang, Chao Xie, Chenning Xiao, Zhiwei Chen, Haibo Parallel Processing Institute Fudan University China

ISBN: (数字)9783642241512

ISBN: (纸本)9783642241505

MapReduce has been shown to be a simple and efficient way to harness the massive resources of clusters. Recently, researchers propose using partitioned global address space (PGAS) based language and runtime to ease the programming of large-scale clusters. In this paper, we present an empirical study on the effectiveness of running MapReduce applications on a typical PGAS language runtime called X10. By tuning the performance of two applications on X10 platforms, we successfully eliminate several performance bottlenecks related to I/O processing. We also identify several remaining problems and propose several approaches to remedying them. Our final performance evaluation on a small-scale multicore cluster shows that the MapReduce applications written with X10 notably outperform those in Hadoop in most cases. Detailed analysis reveals that the major performance advantages come from a simplified task management and data storage scheme. © 2011 Springer-Verlag.

关键词： MapReduce

来源：评论

学校读者我要写书评

暂无评论

A parallel analysis on Scale Invariant Feature Transform (SIFT) algorithm 1

引用

Lecture Notes in Computer Science

作者： Yang, Donglei Liu, Lili Zhu, Feiwen Zhang, Weihua Parallel Processing Institute Fudan University China

ISBN: (数字)9783642241512

ISBN: (纸本)9783642241505

With explosive growth of multimedia data on internet, the effective information retrieval from a large scale of multimedia data becomes more and more important. To retrieve these multimedia data automatically, some features in them must be extracted. Hence, image feature extraction algorithms have been a fundamental component of multimedia retrieval. Among these algorithms, Scale Invariant Feature Transform (SIFT) has been proven to be one of the most robust image feature extraction algorithm. However, SIFT algorithm is not only data intensive but also computation intensive. It takes about four seconds to process an image or a video frame on a general-purpose CPU, which is far from real-time processing requirement. Therefore, accelerating SIFT algorithm is urgently needed. As multi-core CPU becomes more and more popular in recent years, it is natural to employ computing power of multi-core CPU to accelerate SIFT. How to parallelize SIFT to take full use of multi-core capabilities becomes one of the core issues. This paper analyzes available parallelism in SIFT and implements various parallel SIFT algorithms to evaluate which is the most suitable for multi-core system. The final result shows that our parallel SIFT achieves a speedup of 10.46X on 16-core machine. © 2011 Springer-Verlag.

关键词： Computing power

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：