检索结果-内蒙古大学图书馆

26th International Conference on High Performance Computing, Data and Analytics (HiPCW)

作者： Maronas, Marcos Sala, Kevin Mateo, Sergi Ayguade, Eduard Beltran, Vicenc Barcelona Supercomp Ctr BSC Barcelona Spain Univ Politecn Catalunya UPC Barcelona Supercomp Ctr BSC Barcelona Spain

ISBN: (纸本)9781728145358

Shared memory programming models usually provide worksharing and task constructs. The former relies on the efficient fork-join execution model to exploit structured parallelism;while the latter relies on fine-grained synchronization among tasks and a flexible data-flow execution model to exploit dynamic, irregular, and nested parallelism. On applications that show both structured and unstructured parallelism, both worksharing and task constructs can be combined. However, it is difficult to mix both execution models without penalizing the data-flow execution model. Hence, on many applications structured parallelism is also exploited using tasks to leverage the full benefits of a pure data-flow execution model. However, task creation and management might introduce a non-negligible overhead that prevents the efficient exploitation of fine-grained structured parallelism, especially on many-core processors. In this work, we propose worksharing tasks. These are tasks that internally leverage worksharing techniques to exploit fine-grained structured loop-based parallelism. The evaluation shows promising results on several benchmarks and platforms.

关键词： fine grained loop parallelism programming models runtime systems

来源：评论

学校读者我要写书评

暂无评论

Preparation and Optimization of a Diverse Workload for a Large-Scale Heterogeneous System 19

Preparation and Optimization of a Diverse Workload for a Lar...

引用

International Conference for High Performance Computing, Networking, Storage and Analysis (SC)

作者： Karlin, Ian Park, Yoonho de Supinski, Bronis R. Wang, Peng Still, Bert Beckingsale, David Blake, Robert Chen, Tong Cong, Guojing Costa, Carlos Dahm, Johann Domeniconi, Giacomo Epperly, Thomas Fisher, Aaron Schumacher, Sara Kokkila Langer, Steven Le, Hai Lee, Eun Kyung Maruyama, Naoya Que, Xinyu Richards, David Sjogreen, Bjorn Wong, Jonathan Woodward, Carol Yang, Ulrike Zhang, Xiaohua Anderson, Bob Appelhans, David Barnes, Levi Barnes, Peter Bastea, Sorin Boehme, David Bramwell, Jamie A. Brase, Jim Brunheroto, Jose Chen, Barry Cooper, Charway R. DeGroot, Tony Falgout, Rob Gamblin, Todd Gardner, David Glosli, James Gunnels, John Katz, Max Kolev, Tzanio Kuo, I-Feng W. Legendre, Matthew P. Li, Ruipeng Lin, Pei-Hung Lockhart, Shelby McCandless, Kathleen Misale, Claudia Moreno, Jaime Neely, Rob Nelson, Jarom Nimmakayala, Rao O'Brien, Kathryn O'Brien, Kevin Pankajakshan, Ramesh Pearce, Roger Peles, Slaven Regier, Phil Rennich, Steve Schulz, Martin Scott, Howard Sexton, James Shoga, Kathleen Sundram, Shiv Thomas-Collignon, Guillaume Van Essen, Brian Voronin, Alexey Walkup, Bob Wang, Lu Ward, Chris Wen, Hui-Fang White, Dan Young, Christopher Zeller, Cyril Zywicz, Ed Lawrence Livermore Natl Lab Livermore CA 94550 USA IBM TJ Watson Res Ctr Yorktown Hts NY USA NVIDIA Corp Santa Clara CA USA Univ Illinois Champaign IL USA Tech Univ Munich Munich Germany Pacific Northwest Natl Lab Richland WA 99352 USA Sandia Natl Labs Livermore CA 94550 USA

ISBN: (数字)9781450362290

ISBN: (纸本)9781450362290

Productivity from day one on supercomputers that leverage new technologies requires significant preparation. An institution that procures a novel system architecture often lacks sufficient institutional knowledge and skills to prepare for it. Thus, the "Center of Excellence" (CoE) concept has emerged to prepare for systems such as Summit and Sierra, currently the top two systems in the Top 500. This paper documents CoE experiences that prepared a workload of diverse applications and math libraries for a heterogeneous *** describe our approach to this preparation, including our management and execution strategies, and detail our experiences with and reasons for using different programming approaches. Our early science and performance results show that the project enabled significant early seismic science with up to a 14X throughput increase over Cori. In addition to our successes, we discuss our challenges and failures so others may benefit from our experience.

关键词： performance large-scale applications heterogeneous systems GPUs programming models project management

来源：评论

学校读者我要写书评

暂无评论

What's their story? Students share their experiences and understandings of barriers for success within alternative programs designed for "at-risk" youth

What's their story? Students share their experiences and und...

引用

作者： Cooper, Kathryn Jill University of Manitoba

学位级别：硕士

Despite many alternative programs designed for students who have been labelled as unsuccessful within traditional environments, in the urban context where this study takes place, large numbers of students are leaving before receiving high school diplomas. Studies within the educational literature describe the reasons why students are pushed out of schools as well as possible models for alternatives. What is lacking in the research are the voices of students themselves. In response to this gap, questions were posed to four former students that aimed to understand their experiences in order to influence design and programming for future alternative programs. In this context, "alternative" refers to those educational sites attended by students who have typically been failed by more traditional schools. This study, while influenced by narrative inquiry's use of story, uses qualitative semi structured interviews as a methodology. The analysis of these stories raises questions of power, agency, and the role of schools in meeting the needs of all learners. The participants' stories show a need for alternative programs to have clearly defined and articulated visions, attention to the overall purpose and intent of course activities and learning experiences, as well as a more holistic view of the student. Reimagining equitable student-teacher relationships, open access for different populations of learners, advocacy for students, and design models that are constantly re-examined and upgraded based on the needs of students were important conclusions. Despite the initial intention to focus this study on alternative schools and programs for so called "at-risk" learners, the findings suggest that educators and policy makers should pay attention to student experience within alternative education in order to improve those programs as well as their traditional counterparts so there is no longer a need for alternatives.

关键词： Alternative Education Marginalized Students programming models

来源：评论

学校读者我要写书评

暂无评论

Safe, Secure Executions at the Network Edge Coordinating Cloud, Edge, and Fog Computing

引用

IEEE SOFTWARE 2018年第1期35卷 30-37页

作者： Makitalo, Niko Ometov, Aleksandr Kannisto, Joona Andreev, Sergey Koucheryavy, Yevgeni Mikkonen, Tommi Univ Helsinki Dept Comp Sci Helsinki Finland Tampere Univ Technol Lab Elect & Commun Engn Tampere Finland Tampere Univ Technol Lab Pervas Comp Tampere Finland Univ Helsinki Software Engn Helsinki Finland

System design where cyber-physical applications are securely coordinated from the cloud may simplify the development process. However, all private data are then pushed to these remote “swamps,” and human users lose actual control as compared to when the applications are executed directly on their devices. At the same time, computing at the network edge is still lacking support for such straightforward multidevice development, which is essential for a wide range of dynamic cyber-physical services. This article proposes a novel programming model as well as contributes the associated secure-connectivity framework for leveraging safe coordinated device proximity as an additional degree of freedom between the remote cloud and the safety-critical network edge, especially under uncertain environment constraints. This article is part of a special issue on Software Safety and Security Risk Mitigation in Cyber-physical Systems.

关键词： Cloud Computing Data Privacy Mobile Computing Security Of Data Secure Executions Fog Computing Cyber Physical Applications Private Data Dynamic Cyber Physical Services Secure Connectivity Framework Safety Critical Network Edge Security Risk Mitigation Edge Computing Cyber Physical Systems Cloud Computing Edge Computing programming Computer Security Software Development Cyber Physical Systems Computational Modeling Fog Computing Edge Computing programming models Security And Trust Proximate Connectivity Systems And Software Design Programmable World Software Development Software Engineering

来源：评论

学校读者我要写书评

暂无评论

Task-based programming in COMPSs to converge from HPC to big data

引用

INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS 2018年第1期32卷 45-60页

作者： Conejero, Javier Corella, Sandra Badia, Rosa M. Labarta, Jesus BSC C Jordi Girona 31 Barcelona 08034 Spain BSC Work Flows & Distributed Comp Res Grp Barcelona Spain BSC Comp Sci Res Dept Barcelona Spain CSIC IIIA Barcelona Spain

Task-based programming has proven to be a suitable model for high-performance computing (HPC) applications. Different implementations have been good demonstrators of this fact and have promoted the acceptance of task-based programming in the OpenMP standard. Furthermore, in recent years, Apache Spark has gained wide popularity in business and research environments as a programming model for addressing emerging big data problems. COMP Superscalar (COMPSs) is a task-based environment that tackles distributed computing (including Clouds) and is a good alternative for a task-based programming model for big data applications. This article describes why we consider that task-based programming models are a good approach for big data applications. The article includes a comparison of Spark and COMPSs in terms of architecture, programming model, and performance. It focuses on the differences that both frameworks have in structural terms, on their programmability interface, and in terms of their efficiency by means of three widely known benchmarking kernels: Wordcount, Kmeans, and Terasort. These kernels enable the evaluation of the more important functionalities of both programming models and analyze different work flows and conditions. The main results achieved from this comparison are (1) COMPSs is able to extract the inherent parallelism from the user code with minimal coding effort as opposed to Spark, which requires the existing algorithms to be adapted and rewritten by explicitly using their predefined functions, (2) it is an improvement in terms of performance when compared with Spark, and (3) COMPSs has shown to scale better than Spark in most cases. Finally, we discuss the advantages and disadvantages of both frameworks, highlighting the differences that make them unique, thereby helping to choose the right framework for each particular objective.

关键词： programming models distributed computing framework comparison Big Data programming

来源：评论

学校读者我要写书评

暂无评论

Evolution of a minimal parallel programming model

引用

INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS 2018年第1期32卷 4-13页

作者： Lusk, Ewing Butler, Ralph Pieper, Steven C. Argonne Natl Lab Math & Comp Sci Div 9700 S Cass Ave Argonne IL 60439 USA Argonne Natl Lab Div Phys 9700 S Cass Ave Argonne IL 60439 USA Middle Tennessee State Univ Comp Sci Murfreesboro TN 37130 USA

We take a historical approach to our presentation of self-scheduled task parallelism, a programming model with its origins in early irregular and nondeterministic computations encountered in automated theorem proving and logic programming. We show how an extremely simple task model has evolved into a system, asynchronous dynamic load balancing (ADLB), and a scalable implementation capable of supporting sophisticated applications on today's (and tomorrow's) largest supercomputers;and we illustrate the use of ADLB with a Green's function Monte Carlo application, a modern, mature nuclear physics code in production use. Our lesson is that by surrendering a certain amount of generality and thus applicability, a minimal programming model (in terms of its basic concepts and the size of its application programmer interface) can achieve extreme scalability without introducing complexity.

关键词： Parallel computing programming models load balancing automated theorem proving nuclear physics

来源：评论

学校读者我要写书评

暂无评论

RedThreads: An Interface for Application-Level Fault Detection/Correction Through Adaptive Redundant Multithreading

引用

INTERNATIONAL JOURNAL OF PARALLEL programming 2018年第2期46卷 225-251页

作者： Hukerikar, Saurabh Teranishi, Keita Diniz, Pedro C. Lucas, Robert F. Univ Southern Calif Inst Informat Sci 4676 Admiralty WaySuite 1001 Marina Del Rey CA 90292 USA Sandia Natl Labs 7011 East Ave Livermore CA 94551 USA

In the presence of accelerated fault rates, which are projected to be the norm on future exascale systems, it will become increasingly difficult for high-performance computing (HPC) applications to accomplish useful computation. Due to the fault-oblivious nature of current HPC programming paradigms and execution environments, HPC applications are insufficiently equipped to deal with errors. We believe that HPC applications should be enabled with capabilities to actively search for and correct errors in their computations. The redundant multithreading (RMT) approach offers lightweight replicated execution streams of program instructions within the context of a single application process. However, the use of complete redundancy incurs significant overhead to the application performance. In this paper we present RedThreads, an interface that provides application-level fault detection and correction based on RMT, but applies the thread-level redundancy adaptively. We describe the RedThreads syntax and semantics, and the supporting compiler infrastructure and runtime system. Our approach enables application programmers to scope the extent of redundant computation. Additionally, the runtime system permits the use of RMT to be dynamically enabled, or disabled, based on the resiliency needs of the application and the state of the system. Our experimental results demonstrate how adaptive RMT exploits programmer insight and runtime inference to dynamically navigate the trade-off space between an application's resilience coverage and the associated performance overhead of redundant computation.

关键词： Resilience Exascale Redundant multithreading programming models Runtime systems Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

A lightweight approach to performance portability with targetDP

引用

INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS 2018年第2期32卷 288-301页

作者： Gray, Alan Stratford, Kevin Univ Edinburgh EPCC Edinburgh EH9 3FD Midlothian Scotland

Leading high performance computing systems achieve their status through use of highly parallel devices such as NVIDIA graphics processing units or Intel Xeon Phi many-core CPUs. The concept of performance portability across such architectures, as well as traditional CPUs, is vital for the application programmer. In this paper we describe targetDP, a lightweight abstraction layer which allows grid-based applications to target data parallel hardware in a platform agnostic manner. We demonstrate the effectiveness of our pragmatic approach by presenting performance results for a complex fluid application (with which the model was co-designed), plus separate lattice quantum chromodynamics particle physics code. For each application, a single source code base is seen to achieve portable performance, as assessed within the context of the Roofline model. TargetDP can be combined with Message Passing Interface (MPI) to allow use on systems containing multiple nodes: we demonstrate this through provision of scaling results on traditional and graphics processing unit-accelerated large scale supercomputers.

关键词： Performance portability programming models high performance computing graphics processing unit manycore computational physics

来源：评论

学校读者我要写书评

暂无评论

Using approximate dynamic programming to model military force mix adaptation

引用

Military Operations Research 2018年第4期23卷 25-36页

作者： Southerland, Jason Loerch, Andrew Center for Army Analysis United States George Mason University United States

来源：评论

学校读者我要写书评

暂无评论

AIRA: A Framework for Flexible Compute Kernel Execution in Heterogeneous Platforms

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2018年第2期29卷 269-282页

作者： Lyerly, Robert Murray, Alastair Barbalace, Antonio Ravindran, Binoy Virginia Tech Dept Elect & Comp Engn Blacksburg VA 24061 USA

Heterogeneous-ISA computing platforms have become ubiquitous, and will be used for diverse workloads which render static mappings of computation to processors inadequate. Dynamic mappings which adjust an application's usage in consideration of platform workload can reduce application latency and increase throughput for heterogeneous platforms. We introduce AIRA, a compiler and runtime for flexible execution of applications in CPU-GPU platforms. Using AIRA, we demonstrate up to a 3.78x speedup in benchmarks from Rodinia and Parboil, run with various workloads on a server-class platform. Additionally, AIRA is able to extract up to an 87 percent increase in platform throughput over a static mapping.

关键词： Heterogeneous architectures compilers runtimes programming models

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：