检索结果-内蒙古大学图书馆

IEEE High Performance Extreme computing Conference (HPEC)

作者： Barnell, Mark Raymond, Courtney Brown, Daniel Wilson, Matthew Cote, Eric Air Force Res Lab Informat Directorate Rome NY 13441 USA SRC Inc Radars & Sensors Div N Syracuse NY USA

ISBN: (纸本)9781728150208

For the first time ever, advanced machine learning (ML) compute architectures, techniques, and methods were demonstrated on United States Geological Survey (USGS) optical imagery and Department of Defense (DoD) Synthetic Aperture Radar (SAR) imagery, simultaneously, using IBM's new NS16e neurosynaptic processor board comprised of 16 TrueNorth chips. The Air Force Research Laboratory (AFRL) Information Directorate Advanced computing and Communications Division continues to develop and demonstrate new bio-inspired computing algorithms and architectures, designed to provide advanced, ultra-low power, ground and airborne High-Performance computing (HPC) solutions to meet operational and tactical, real-time processing needs for Intelligence, Surveillance, and Reconnaissance (ISR) missions on small form factor hardware, and in Size, Weight and Power (SWaP) constrained environments. With an average throughput of 16,000 inferences per second, the system provided a processing efficiency of 1,066 inferences per Watt. The NS16e power utilization never exceeded 15 Watts for this application. The contribution of power consumption from TrueNorth processors was bound to less than 5.5 Watts.

关键词： computing technologies for challenging form factors radar processing and exploitation data intensive computing Machine Learning (ML) High Performance computing (HPC) Artificial Intelligence (AI) region proposal networks object detection Deep Learning (DL) Convolutional Neural Networks (CNN) advanced processor architectures autonomy Electro-Optical (EO) Intelligence Surveillance and Reconnaissance (ISR) neuromorphic computing

来源：评论

学校读者我要写书评

暂无评论

Hybrid Resource Management for HPC and data intensive Workloads 19

Hybrid Resource Management for HPC and Data Intensive Worklo...

引用

19th Annual IEEE/ACM International Symposium on Cluster, Cloud, and Grid computing (CCGRID)

作者： Souza, Abel Rezaei, Mohamad Laure, Erwin Tordsson, Johan Umea Univ Dept Comp Sci Umea Sweden KTH Royal Inst Technol PDC Ctr High Performance Comp Stockholm Sweden

ISBN: (纸本)9781728109121

High Performance computing (HPC) and data intensive (DI) workloads have been executed on separate clusters using different tools for resource and application management. With increasing convergence, where modern applications are composed of both types of jobs in complex workflows, this separation becomes a growing overhead and the need for a common platform increases. Executing both workload classes on the same clusters not only enables hybrid workflows, but can also increase system efficiency, as available hardware often is not fully utilized by applications. While HPC systems are typically managed in a coarse grained fashion, with exclusive resource allocations, DI systems employ a finer grained regime, enabling dynamic allocation and control based on application needs. On the path to full convergence, a useful and less intrusive step is a hybrid resource management system allowing the execution of DI applications on top of standard HPC scheduling systems. In this paper we present the architecture of a hybrid system enabling dual-level scheduling for DI jobs in HPC infrastructures. Our system takes advantage of real-time resource profiling to efficiently co-schedule HPC and DI applications. The architecture is easily extensible to current and new types of distributed applications, allowing efficient combination of hybrid workloads on HPC resources with increased job throughput and higher overall resource utilization. The implementation is based on the Sturm and Mesos resource managers for HPC and DI jobs. Experimental evaluations in a real cluster based on a set of representative HPC and DI applications demonstrate that our hybrid architecture improves resource utilization by 20%, with 12% decrease on queue makespan while still meeting all deadlines for HPC jobs.

关键词： Resource Management High Performance computing data intensive computing Mesos Sturm Bootstrapping

来源：评论

学校读者我要写书评

暂无评论

A data intensive heuristic approach to the two-stage streaming scheduling problem

引用

JOURNAL OF COMPUTER AND SYSTEM SCIENCES 2017年 89卷 64-79页

作者： Liang, Wei Hu, Chunhua Wu, Min Jin, Qun Hunan Univ Commerce Key Lab Hunan Prov Mobile Business Intelligence Changsha Hunan Peoples R China Hunan Univ Commerce Mobile E Busines Collaborat Innovat Ctr Hunan Pro Changsha Hunan Peoples R China China Univ Geosci Sch Automat Wuhan Hubei Peoples R China Waseda Univ Fac Human Sci Tokorozawa Saitama Japan

data intensive computing (DIC) provides a high performance computing approach to process large volume of data. In this study, a new formalization is introduced to present the two-stage DIC task execution in a stream manner. A novel heuristic algorithm is proposed for the scheduling problem due to the NP complexity. The theoretical approximation ratio bounds for the heuristic are analyzed and confirmed by the experimental evaluation. Overall, we observe that the proposed method conducts average 1.2 times makespan than the theoretic bound of the optimal solution. Besides, the proposed method outperforms the GA and FIFO scheduling schemes with overall improvements. (C) 2017 Elsevier Inc. All rights reserved.

关键词： data intensive computing Scheduling Makespan NP-hard

来源：评论

学校读者我要写书评

暂无评论

Modeling of large-scale social network services based on mechanisms of information diffusion: Sina Weibo as a case study

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 2017年 74卷 291-301页

作者： Wang, Ru Rho, Seungmin Chen, Bo-Wei Cai, Wandong Northwestern Polytech Univ Sch Comp Sci & Engn Xian 710072 Peoples R China Sungkyul Univ Dept Media Software Anyang South Korea Monash Univ Malaysia Campus Sch Informat Technol Subang Jaya Selangor Malaysia

The purpose of this paper is to investigate the characteristics of the dissemination of information in the community. A variety of possible factors that affect the dissemination of information in Sina Weibo have been discussed. By analyzing the process of the information dissemination in the community of Sina Weibo, we found the information dissemination of Weibo community and the dynamic model are very similar. With the aid of data intensive computing theory, the various features have been mined and modeled. The dynamic model is improved and redefined to characterize the community. Then the SEINR model is proposed. The basic reproductive number, the existence of equilibrium point and the stability of the network are analyzed and proved in detail. By comparing with real data in Weibo community, we show that the SEINR model accurately reflects the dissemination of information community. Furthermore, we investigate the SEINR model in detail to show the influences of different parameters on information dissemination by simulations. (C) 2016 Elsevier B.V. All rights reserved.

关键词： Sina Weibo data intensive computing SEINR Propagation dynamics Basic reproduction number Equilibrium point

来源：评论

学校读者我要写书评

暂无评论

Goal-based composition of scalable hybrid analytics for heterogeneous architectures

引用

JOURNAL OF PARALLEL AND DISTRIBUTED computing 2017年 108卷 59-73页

作者： Coetzee, P. Jarvis, S. A. Univ Warwick Dept Comp Sci Coventry W Midlands England British Lib Alan Turing Inst London England

Crafting scalable analytics in order to extract actionable business intelligence is a challenging endeavour, requiring multiple layers of expertise and experience. Often, this expertise is irreconcilably split between an organisation's engineers and subject matter domain experts. Previous approaches to this problem have relied on technically adept users with tool-specific training. Such an approach has a number of challenges: Expertise - There are few data-analytic subject domain experts with in-depth technical knowledge of compute architectures;Performance - Analysts do not generally make full use of the performance and scalability capabilities of the underlying architectures;Heterogeneity - calculating the most performant and scalable mix of real-time (on-line) and batch (off-line) analytics in a problem domain is difficult;Tools - Supporting frameworks will often direct several tasks, including, composition, planning, code generation, validation, performance tuning and analysis, but do not typically provide end-to-end solutions embedding all of these activities. In this paper, we present a novel semi-automated approach to the composition, planning, code generation and performance tuning of scalable hybrid analytics, using a semantically rich type system which requires little programming expertise from the user. This approach is the first of its kind to permit domain experts with little or no technical expertise to assemble complex and scalable analytics, for hybrid on- and off-line analytic environments, with no additional requirement for low-level engineering support. This paper describes (i) an abstract model of analytic assembly and execution, (ii) goal-based planning and (iii) code generation for hybrid on- and off-line analytics. An implementation, through a system which we call MENDELEEV, is used to (iv) demonstrate the applicability of this technique through a series of case studies, where a single interface is used to create analytics that can be run simu

关键词： data science Hybrid analytics Analytic planning Streaming analysis Hadoop data intensive computing Heterogeneous compute

来源：评论

学校读者我要写书评

暂无评论

Online Scheduling and Interference Alleviation for Low-Latency, High-Throughput Processing of data Streams

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2017年第12期28卷 3553-3569页

作者： Buddhika, Thilina Stern, Ryan Lindburg, Kira Ericson, Kathleen Pallickara, Shrideep Colorado State Univ Deptartment Comp Sci Ft Collins CO 80523 USA

data Streams occur naturally in several observational settings and often need to be processed with a low latency. Streams pose unique challenges: they have no preset lifetimes, the traffic on these streams may be bursty, and data arrival rates on these streams can be quite high. Furthermore, stream processing computations are generally stateful where the outcome of processing a data stream packet depends on the state that builds up within the computation over multiple, successive rounds of execution. As the number of streams increases, stream processing computations need to be orchestrated over a collection of machines. Achieving timeliness and high throughput in such settings is a challenge. Optimal scheduling of stream processing computations is an instance of the resource constrained scheduling problem, and depending on the precise formulation of the problem can be characterized as either NP-Complete or NP-Hard. We have designed an algorithm for online scheduling of stream processing computations. Our algorithm focuses on reducing interference that adversely impacts performance of stream processing computations. Our measure of interference is based on stream packet arrivals at a particular machine, the accompanying resource utilization encompassing CPU, memory and network utilization, and the resource utilization at machines comprising the cluster. Our algorithm performs continuous, incremental detection of interference experienced by computations and performing migrations to alleviate them.

关键词： Low-latency stream processing online scheduling data intensive computing

来源：评论

学校读者我要写书评

暂无评论

A Framework for Multitasking data-intensive Management Services in High Performance computing Environments 1

A Framework for Multitasking Data-Intensive Management Servi...

引用

1st IEEE International Conference on Big data computing Service and Applications (BigdataService)

作者： Kulasekaran, Sivakumar Esteva, Maria Trelogan, Jessica Liu, Si Texas Adv Comp Ctr Houston TX 77054 USA Inst Class Archaeol London England Univ Texas Austin Austin TX 78712 USA

ISBN: (纸本)9781479981281

data management entails a continuum of tasks to develop sustainable and reusable collections throughout their lifecycle. Large collections with complex data formats and structures may require what we define as "multitasking data management," involving a combination of manual and automated iterative tasks. When conducted in a desktop computing environment by curators, these tasks can be labor-intensive and disruptive of research. While the process can be made much more efficient within a data-intensive High Performance computing (DIC/HPC) infrastructure, it remains a challenge to implement generalizable services so that automated workflows can be easily performed by non-expert users. This paper introduces a framework for automating data management activities as data-intensive computing jobs within a multitasking workflow. Using as a case study a set of legacy data from an archaeological collection in need of reorganization, we identified the steps required to re-sort and move approximately 27,000 data files into a structured collection architecture. Because not all data management workflows are the same, and because there are a wide range of requirements for job submission within data-intensive HPC resources, we derived a set of generalizable modules that can be used as a guide for curators and HPC consultants. This framework may accommodate collections with different data types and data management requirements and can be conducted by curators trained in HPC usage but without ample computational expertise. Upon testing, we implemented the framework as a service on a DIC/HPC cluster.

关键词： Multitasking data management services high performance computing archaeology data data intensive computing

来源：评论

学校读者我要写书评

暂无评论

Pressure-Driven Hardware Managed Thread Concurrency for Irregular Applications 7

Pressure-Driven Hardware Managed Thread Concurrency for Irre...

引用

7th Workshop on Irregular Applications - Architectures and Algorithms (IA3)

作者： Leidel, John D. Wang, Xi Chen, Yong Texas Tech Univ Box 43104 Lubbock TX 79409 USA

ISBN: (纸本)9781450351362

Given the increasing importance of efficient data intensive computing, we find that modern processor designs are not well suited to the irregular memory access patterns found in these algorithms. This research focuses on mapping the compiler's instruction cost scheduling logic to hardware managed concurrency controls in order to minimize pipeline stalls. In this manner, the hardware modules managing the low-latency thread concurrency can be directly understood by modern compilers. We introduce a thread context switching method that is managed directly via a set of hardware-based mechanisms that are coupled to the compiler instruction scheduler. As individual instructions from a thread execute, their respective cost is accumulated into a control register. Once the register reaches a pre-determined saturation point, the thread is forced to context switch. We evaluate the performance benefits of our approach using a series of 24 benchmarks that exhibit performance acceleration of up to 14.6X.

关键词： data intensive computing irregular algorithms thread concurrency context switching

来源：评论

学校读者我要写书评

暂无评论

Performance analysis of data intensive cloud systems based on data management and replication: a survey

引用

DISTRIBUTED AND PARALLEL dataBASES 2016年第2期34卷 179-215页

作者： Malik, Saif Ur Rehman Khan, Samee U. Ewen, Sam J. Tziritas, Nikos Kolodziej, Joanna Zomaya, Albert Y. Madani, Sajjad A. Min-Allah, Nasro Wang, Lizhe Xu, Cheng-Zhong Malluhi, Qutaibah Marwan Pecero, Johnatan E. Balaji, Pavan Vishnu, Abhinav Ranjan, Rajiv Zeadally, Sherali Li, Hongxiang COMSATS Inst Informat Technol Islamabad Pakistan N Dakota State Univ Fargo ND 58105 USA Shenzhen Inst Adv Technol Shenzhen Guangdong Peoples R China Univ Bielsko Biala Bielsko Biala Poland Univ Sydney Sydney NSW 2006 Australia Univ Dammam Dammam Saudi Arabia Chinese Acad Sci Beijing Peoples R China Wayne State Univ Detroit MI USA Qatar Univ Doha Qatar Univ Luxembourg Walferdange Luxembourg Argonne Natl Lab Lemont IL USA Pacific NW Natl Lab Richland WA 99352 USA CSIRO ICT Ctr Marsfield NSW Australia Univ Dist Columbia Washington DC 20008 USA Univ Louisville Louisville KY 40292 USA

As we delve deeper into the 'Digital Age', we witness an explosive growth in the volume, velocity, and variety of the data available on the Internet. For example, in 2012 about 2.5 quintillion bytes of data was created on a daily basis that originated from myriad of sources and applications including mobile devices, sensors, individual archives, social networks, Internet of Things, enterprises, cameras, software logs, etc. Such 'data Explosions' has led to one of the most challenging research issues of the current Information and Communication Technology era: how to optimally manage (e.g., store, replicated, filter, and the like) such large amount of data and identify new ways to analyze large amounts of data for unlocking information. It is clear that such large data streams cannot be managed by setting up on-premises enterprise database systems as it leads to a large up-front cost in buying and administering the hardware and software systems. Therefore, next generation data management systems must be deployed on cloud. The cloud computing paradigm provides scalable and elastic resources, such as data and services accessible over the Internet Every Cloud Service Provider must assure that data is efficiently processed and distributed in a way that does not compromise end-users' Quality of Service (QoS) in terms of data availability, data search delay, data analysis delay, and the like. In the aforementioned perspective, data replication is used in the cloud for improving the performance (e.g., read and write delay) of applications that access data. Through replication a data intensive application or system can achieve high availability, better fault tolerance, and data recovery. In this paper, we survey data management and replication approaches (from 2007 to 2011) that are developed by both industrial and research communities. The focus of the survey is to discuss and characterize the existing approaches of data replication and management that tackle the resource u

关键词： Replication data management Cloud computing systems Performance gradation data intensive computing

来源：评论

学校读者我要写书评

暂无评论

Cross-Institutional Research Cyberinfrastructure for data intensive Science

Cross-Institutional Research Cyberinfrastructure for Data In...

引用

IEEE High Performance Extreme computing Conference (HPEC)

作者： Lenhardt, W. Christopher Conway, Mike Scott, Erik Blanton, Brian Krishnamurthy, Ashok Hadzikadic, Mirsad Vouk, Mladen Wilson, Alyson Univ N Carolina Renaissance Comp Inst RENCI Chapel Hill NC 27514 USA Univ N Carolina Charlotte NC 28223 USA North Carolina State Univ Raleigh NC 27695 USA

ISBN: (纸本)9781509035250

This paper describes a multi-institution effort to develop a "data science as a service" platform. This platform integrates advanced federated data management for small to large datasets, access to high performance computing, distributed computing and advanced networking. The goal is to develop a platform that is flexible and extensible while still supporting domain research and avoiding the walled garden problem. Some preliminary lessons learned and next steps will also be outlined.

关键词： distributed computing risk analytics distributed data open system architectures data intensive computing big data and distributed computing

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：