检索结果-内蒙古大学图书馆

Analyzing and optimizing the performance and energy efficiency of transactional scientific applications on large-scale NUMA systems with HTM support

引用

JOURNAL OF parallel And distributed COMPUTING 2019年 127卷 1-17页

作者： Park, Jinsu Baek, Woongki UNIST Sch ECE 50 UNIST Gil Ulsan South Korea

Hardware transactional memory (HTM) is widely supported by commodity processors. While the effectiveness of HTM has been evaluated based on small-scale multi-core systems, it still remains unexplored to quantify the performance and energy efficiency of HTM for scientific workloads on large-scale NUMA systems, which have been increasingly adopted to high-performance computing. To bridge this gap, this work investigates the performance and energy-efficiency impact of HTM on scientific applications on large-scale NUMA systems. Specifically, we quantify the performance and energy efficiency of HTM for scientific workloads based on the widely-used CLOMP-TM benchmark. We then discuss a set of generic software optimizations, which effectively improve the performance and energy efficiency of transactional scientific workloads on large-scale NUMA systems. Further, we present case studies in which we apply a set of the performance and energy-efficiency optimizations to representative transactional scientific applications and investigate the potential for high-performance and energy-efficient runtime support. (C) 2019 Elsevier Inc. All rights reserved.

关键词： Hardware transactional memory Non-uniform memory access Scientific applications High performance Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

Rumor Spreading with Cross Propagation in Multilayer Social Networks 17

Rumor Spreading with Cross Propagation in Multilayer Social ...

引用

IEEE Int Conf on parallel and distributed processing with applications, Big Data and Cloud Computing, Sustainable Computing and Communications, Social Computing and Networking (ISPA/BDCloud/SocialCom/SustainCom)

作者： Han, Qiyi Gu, Musong You, Lei Miao, Fang Chengdu Univ Sch Informat Sci & Engn 2025 Ave Been Chengdu 610106 Sichuan Peoples R China

ISBN: (纸本)9781728143286

Social networks have achieved great success by making social interaction online. Malicious users, whereas, make use of social networks to spread rumors. Recent researches manifest that the union of social applications can reach better efficiency. Unfortunately, novel security threats are emerging when malicious users take advantage of this opportunity to spread rumors from a single social network to multilayer networks. In this paper, concerning the cross propagation in the multilayer social networks, the S2IR2 model is therefore proposed to analyze the rumor spreading. The rumor spreading process is molded by mean-field equations. With the simulations of various parameters, the numerical results give the performances of propagation speed and range. For restraining the rumor spreading, the security strategies are concluded.

关键词： rumor spreading multilayer networks social networks S2IR2 model

来源：评论

学校读者我要写书评

暂无评论

parallel Computation of Division over GF(2ⁿ) Covering Divide-by-zero Based on Tile Assembly Model 17

Parallel Computation of Division over GF(2<SUP>n</SUP>) Cove...

引用

作者： Li, Yongnan Xiao, Limin Peoples Publ Secur Univ China Sch Natl Secur & Counter Terrorism Beijing 100038 Peoples R China Beihang Univ Sch Comp Sci & Engn Beijing 100191 Peoples R China

ISBN: (纸本)9781728143286

This paper proposes how to compute division over finite field GF(2(n)) that could cover the case of zero divisor based on tile assembly model. Key functions are accomplished by combining n final configurations of two different sub-models. The final configuration of every sub-model contains 2n(2) - n computation tiles which perform 10 different functions. The coding ways of most computation tiles are optimized than the previous model to improve the complexity. The highest bit of modulus number is omitted to simplify the encoding way of computation tiles in the assembly process. This model requires 4591 types of computation tiles and 13 types of boundary tiles. The assembly time complexity is 2n(2) - 1 and the space complexity is n(4).

关键词： Finite field GF(2(n)) Tile assembly model parallel computing Division Divide-by-zero

来源：评论

学校读者我要写书评

暂无评论

MIDAS: Multilinear detection at scale

引用

JOURNAL OF parallel And distributed COMPUTING 2019年 132卷 363-382页

作者： Ekanayake, Saliya Cadena, Jose Wickramasinghe, Udayanga Vullikanti, Anil Lawrence Berkeley Natl Lab Performance & Algorithms Res PAR Grp Berkeley CA USA Lawrence Livermore Natl Lab Computat Engn Div Livermore CA 94550 USA Indiana Univ Dept Comp Sci Bloomington IN 47405 USA Univ Virginia Dept Comp Sci Charlottesville VA 22903 USA Univ Virginia Biocomplex Inst & Initiat Charlottesville VA 22903 USA

We focus on two classes of problems in graph mining: (1) finding trees and (2) anomaly detection in complex networks using scan statistics. These are fundamental problems in a broad class of applications. Most of the parallel algorithms for such problems are either based on heuristics, which do not scale very well, or use techniques like color coding, which have a high memory overhead. In this paper, we develop a novel approach for parallelizing both these classes of problems, using an algebraic representation of subgraphs as monomials-this methodology involves detecting multilinear terms in multivariate polynomials. Our algorithms show good scaling over a large regime, and they run on networks with close to half one billion edges. The resulting parallel algorithm for trees is able to scale to subgraphs of size 18, which has not been done before, and it significantly outperforms the best prior color coding based method (FASCIA) by more than two orders of magnitude. Our algorithm for network scan statistics is the first such parallelization, and it is able to handle a broad class of scan statistics functions with the same approach. Published by Elsevier Inc.

关键词： Subgraph isomorphism distributed graph algorithms Graph scan statistics Multilinear detection Parameterized complexity

来源：评论

学校读者我要写书评

暂无评论

Small Sized File Storage Problems in Hadoop distributed File System 2

Small Sized File Storage Problems in Hadoop Distributed File...

引用

2nd international Conference on Smart Systems and Inventive Technology, ICSSIT 2019

作者： Alange, Neeta Mathur, Anjali Koneru Lakshmaiah Education Foundation Department of Computer Science and Engineering Vaddeswaram AP India

ISBN: (纸本)9781728121192

Hadoop distributed File System (HDFS) is widely used to store the files, which are having heavy size. HDFS is so-called as distributed file system, which intends to store and access the huge volume of data. It is highly fault tolerant system. HDFS opts for the characteristic of cloud repository base, which is having capacity of ready-to-scale, good in performance efficiency and low-cost storing ability. It offers parallel processing therefore, it has high throughput. HDFS is more appropriate for applications, which are having massive datasets. However, it is inefficient for accumulation of multiple small sized files and faces many problems while processing. In this paper, the number of existing techniques have compared in terms of their performance throughput and a comparative chart has been prepared to compare the solutions in a better way for small file problem in Hadoop. By analyzing the chart, a better and efficient solution will able, to propose to store the small sized files in faster way. © 2019 IEEE.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

A Deep Recurrent Neural Network Based Predictive Control Framework for Reliable distributed Stream Data processing 33

A Deep Recurrent Neural Network Based Predictive Control Fra...

引用

33rd IEEE international parallel and distributed processing symposium (IPDPS)

作者： Xu, Jielong Tang, Jian Xu, Zhiyuan Yin, Chengxiang Kwiat, Kevin Kamhoua, Charles Syracuse Univ Dept Elect Engn & Comp Sci Syracuse NY 13244 USA US Air Force Res Lab AFRL Wright Patterson AFB OH USA US Army Res Lab ARL Adelphi MD USA

ISBN: (纸本)9781728112466

In this paper, we present design, implementation and evaluation of a novel predictive control framework to enable reliable distributed stream data processing, which features a Deep Recurrent Neural Network (DRNN) model for performance prediction, and dynamic grouping for flexible control. Specifically, we present a novel DRNN model, which makes accurate performance prediction with careful consideration for interference of co-located worker processes, according to multi-level runtime statistics. Moreover, we design a new grouping method, dynamic grouping, which can distribute/re-distribute data tuples to downstream tasks according to any given split ratio on the fly. So it can be used to re-direct data tuples to bypass misbehaving workers. We implemented the proposed framework based on a widely used distributed Stream Data processing System (DSDPS), Storm. For validation and performance evaluation, we developed two representative stream data processing applications: Windowed URL Count and Continuous Queries. Extensive experimental results show: 1) The proposed DRNN model outperforms widely used baseline solutions, ARIMA and SVR, in terms of prediction accuracy;2) dynamic grouping works as expected;and 3) the proposed framework enhances reliability by offering minor performance degradation with misbehaving workers.

关键词： Deep Learning Recurrent Neural Network distributed Stream Data processing Storm Prediction

来源：评论

学校读者我要写书评

暂无评论

Silicon Proven 1.8 μmx9.2 μm 65-nm Digital Bit Generator for Hardware Security applications

引用

IEEE TRANSACTIONS ON CIRCUITS And SYSTEMS II-EXPRESS BRIEFS 2019年第10期66卷 1713-1717页

作者： Zooker, David Avital, Moshe Weizman, Yoav Fish, Alexander Keren, Osnat Bar Ilan Univ Fac Engn IL-52900 Ramat Gan Israel

Many circuit/gate level countermeasures require randomness to provide immunity against power analysis attacks. These applications usually require up to 64 bit sequences at different locations in the design in parallel. While these sequences are not required to be truly random, they must be independent. In this brief, we present, for the first time, results of silicon proven compact bit generator (cBG). The measurements prove that cBGs are able to amplify random fluctuations in the supply voltage to successfully produce the required bit sequences. The results show that less than 0.00014 random bits (out of 4 bits) are lost because of the negligible dependence between neighbouring generators. Additionally, the generators produce high enough entropy (minimum of 0.9656 over 10 Mbits) so they can be used as distributed generators for gate level countermeasures. The generators were tested with clock frequencies up to 100 MHz and the measured energy per bit of a single generator is 45 pJ@100 MHz.

关键词： Compact bit generator hardware security

来源：评论

学校读者我要写书评

暂无评论

AHEAD: A Tool for Projecting Next-Generation Hardware Enhancements on GPU-Accelerated Systems 33

AHEAD: A Tool for Projecting Next-Generation Hardware Enhanc...

引用

33rd IEEE international parallel and distributed processing symposium (IPDPS)

作者： Abdelhafez, Hazem A. Zimmer, Christopher Vazhkudai, Sudharshan S. Ripeanu, Matei Univ British Columbia Elect & Comp Engn Vancouver BC Canada Oak Ridge Natl Lab Natl Ctr Computat Sci Oak Ridge TN USA

ISBN: (纸本)9781538655559

Starting with the Titan supercomputer (at the Oak Ridge Leadership Computing Facility, OLCF) in 2012, top supercomputers have increasingly leveraged the performance of GPUs to support large-scale computational science. The current No. 1 machine, the 200 petaflop Summit system at OLCF, is a GPU-based machine. Accelerator-based architectures, however, add additional complexity due to node heterogeneity. To inform procurement decisions, supercomputing centers need the tools to quickly model the impact of changes of the node architectures on application performance. We present AHEAD, a profiling and modeling tool to quantify the impact of intra-node communication mechanism (e.g., PCI or NVLink) on application performance. Our experiments show average weighted relative errors of similar to 19% and similar to 23% for five CORAL-2 (a collaboration between multiple US Department of Energy, DOE, labs to procure exascale systems) and 12 Rodinia benchmarks respectively, without running the applications on the target future node.

关键词： Computational modeling Data transfer Graphics processing units Data models Analytical models Tools Kernel

来源：评论

学校读者我要写书评

暂无评论

i²Graph: An Incremental Iterative Computation Model for Large Scale Dynamic Graphs 17

i<SUP>2</SUP>Graph: An Incremental Iterative Computation Mod...

引用

作者： Tang, Zhuo He, Mengsi Yang, Li Fu, Zhongming Hunan Univ Coll Comp Sci & Elect Engn Changsha Peoples R China Changsha Univ Sci & Technol Coll Comp & Commun Engn Changsha Peoples R China

ISBN: (纸本)9781728143286

Due to the rapid changes in graph data sets, the mined information will quickly become obsolete, thus the entire data set needs to be re-computed from the beginning, which will result in the waste of computing time and resources. To reduce the cost of such computations, this paper proposes a model called i(2)Graph to support incremental iterative computation for dynamic graphs. Different from the way of traditional iteration, i(2)Graph executes the graph algorithm by reusing the results of the previous graph and performs computation on parts of the graph that has changed. i(2)Graph contains two components: (1) an incremental iterative computation model to improve the execution efficiency of the iterative graph algorithm;and (2) an incremental update method to accelerate the iterative process within the iterative graph algorithm. It is implemented based on Spark GraphX, a popular parallel and distributed computing framework for large-scale graph processing. Experiment results verify the performance advantages of i(2)Graph model when performing some iterative graph algorithms on the dynamic graph, compared with the traditional iteration.

关键词： dynamic graph iterative graph algorithm Incremental computation Spark GraphX

来源：评论

学校读者我要写书评

暂无评论

Smart DAG Task Scheduling with Efficient Pruning-based MCTS Method 17

Smart DAG Task Scheduling with Efficient Pruning-based MCTS ...

引用

作者： Liu, Kui Wu, Zhiwei Wu, Qing Cheng, Yuxia Hangzhou Dianzi Univ 1158 Baiyang 2 St Hangzhou 310018 Peoples R China

ISBN: (纸本)9781728143286

Efficient task scheduling algorithms are important for achieving high performance in heterogeneous computing systems. The DAG (Directed Acyclic Graph) tasks scheduling problem has been extensively studied. However, many previously proposed algorithms are typically heuristic and based on greedy methods, which can not effectively explore large search space of the problem. In this paper, we proposed a smart DAG task scheduling algorithm with efficient pruning-based Monte Carlo Tree Search (MCTS) method. In the task scheduling process, the algorithm adaptively balances the exploration and exploitation of search policies. To further reduce the search space, a new branch-and-bound based pruning method is proposed that significantly improves the algorithm's efficiency. Our experimental evaluation shows the effectiveness of the proposed algorithm. Under a range of different DAGs and heterogeneous configurations, the pruning-based MCTS scheduling algorithm outperforms HEFT, CPOP and PEFT algorithms in terms of scheduling makespan, while restricts the computing overhead below an acceptable threshold.

关键词： DAG scheduling MCTS task graphs pruning

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：