检索结果-内蒙古大学图书馆

20th acm SIGPLAN symposium on principles and practice of parallel programming, PPoPP 2015

作者： Zhang, Kaiyuan Chen, Rong Chen, Haibo Shanghai Key Laboratory of Scalable Computing and Systems Institute of Parallel and Distributed Systems Shanghai Jiao Tong University China

ISBN: (纸本)9781450332057

Graph-structured analytics has been widely adopted in a number of big data applications such as social computation, web-search and recommendation systems. though much prior research focuses on scaling graph-analytics on distributed environments, the strong desire on performance per core, dollar and joule has generated considerable interests of processing large-scale graphs on a single server-class machine, which may have several terabytes of RAM and 80 or more cores. However, prior graph-analytics systems are largely neutral to NUMA characteristics and thus have suboptimal performance. this paper presents a detailed study of NUMA characteristics and their impact on the efficiency of graph-analytics. Our study uncovers two insights: 1) either random or interleaved allocation of graph data will significantly hamper data locality and parallelism;2) sequential inter-node (i.e., remote) memory accesses have much higher bandwidth than both intra- and inter-node random ones. Based on them, this paper describes Polymer, a NUMA-aware graph-analytics system on multicore with two key design decisions. First, Polymer differentially allocates and places topology data, application-defined data and mutable runtime states of a graph system according to their access patterns to minimize remote accesses. Second, for some remaining random accesses, Polymer carefully converts random remote accesses into sequential remote accesses, by using lightweight replication of vertices across NUMA nodes. To improve load balance and vertex convergence, Polymer is further built with a hierarchical barrier to boost parallelism and locality, an edge-oriented balanced partitioning for skewed graphs, and adaptive data structures according to the proportion of active vertices. A detailed evaluation on an 80-core machine shows that Polymer often outperforms the state-of-the-art single-machine graph-analytics systems, including Ligra, X-Stream and Galois, for a set of popular real-world and synthetic grap

关键词： Random access storage

来源：评论

学校读者我要写书评

暂无评论

Para-functional programming: A paradigm for programming multiprocessor systems 86

Para-functional programming: A paradigm for programming mult...

引用

13th AnnuaJ acm symposium on the principles of programming Languages, POPL 1986

作者： Hudak, Paul Smith, Lauren Yale University Department of Computer Science United States

ISBN: (纸本)089791175X

One of the most important pragmatic advantages of functional languages is that concurrency in a program is implicit-there is no need for special constructs to express parallelism as is required in most conventional languages. Furthermore, it is fairly easy for compilers to automatically determine the concurrency as a step toward decomposing a program for execution on a suitable parallel architecture. Yet it is often the case that one knows precisely the optimal decomposition for execution on a particular machine, and one can never expect a compiler to determine such optimal mappings in all cases. this paper is concerned with ways to allow the programmer to explicitly express this mapping of program to machine, by using annotations that, given a few minor constraints, cannot alter the functional semantics of the program. We show through several deeded examples the expressiveness and conciseness of the resulting " para-functional programming methodology, using an experimental language called ParAlfl based on our ideas. We also give a formal denotational description of the mapping semantics using a notion of execution trees.

关键词： Functional programming

来源：评论

学校读者我要写书评

暂无评论

JParEnt: parallel entropy decoding for JPEG decompression on heterogeneous multicore architectures 7

JParEnt: Parallel entropy decoding for JPEG decompression on...

引用

Euro-Par Conference / 7th International Workshop on programming Models and Applications for Multicores and Manycores (PMAM) held in conjunction with the 21st SIGPLAN symposium on principles and practice of parallel programming (PPoPP)

作者： Sodsong, Wasuwee Jung, Minyoung Park, Jinwoo Burgstaller, Bernd Yonsei Univ Dept Comp Sci Seoul South Korea

ISBN: (纸本)9781450341967

the JPEG format employs Huffman codes to compress the entropy data of an image. Huffman codewords are of variable length, which makes parallel entropy decoding a difficult problem. To determine the start position of a codeword in the bitstream, the previous codeword must be decoded first. We present JParEnt, a new approach to parallel entropy decoding for JPEG decompression on heterogeneous multicores. JParEnt conducts JPEG decompression in two steps: (1)an efficient sequential scan of the entropy data on the CPU to determine the start-positions (boundaries) of coefficient blocks in the bitstream, followed by (2)a parallel entropy decoding step on the graphics processing unit (GPU). the block boundary scan constitutes a reinterpretation of the Huffman-coded entropy data to determine codeword boundaries in the bitstream. We introduce a dynamic workload partitioning scheme to account for GPUs of low compute power relative to the CPU. this configuration has become common with the advent of SoCs with integrated graphics processors (IGPs). We leverage additional parallelism through pipelined execution across CPU and GPU. For systems providing a unified address space between CPU and GPU, we employ zero-copy to completely eliminate the data transfer overhead. Our experimental evaluation of JParEnt was conducted on six heterogeneous multicore systems: one server and two desktops with dedicated GPUs, one desktop with an IGP, and two embedded systems. For a selection of more than 1000JPEG images, JParEnt outperforms the SIMD-implementation of the libjpeg-turbo library by up to a factor of 4.3x, and the previously fastest JPEG decompression method for heterogeneous multicores by up to a factor of 2.2x. JParEnt's entropy data scan consumes 45% of the entropy decoding time of libjpeg-turbo on average. Given this new ratio for the sequential part of JPEG decompression, JParEnt achieves up to97% of the maximum attainable speedup (95% on average). On the IGP-based desktop platform,

关键词： JPEG decoding entropy decoding prefix codes heterogeneous multicores GPU programming

来源：评论

学校读者我要写书评

暂无评论

A Peta-scalable CPU-GPU Algorithm for Global Atmospheric Simulations 13

A Peta-scalable CPU-GPU Algorithm for Global Atmospheric Sim...

引用

18th acm SIGPLAN symposium on principles and practice of parallel programming

作者： Yang, Chao Xue, Wei Fu, Haohuan Gan, Lin Li, Linfeng Xu, Yangtong Lu, Yutong Sun, Jiachang Yang, Guangwen Zheng, Weimin Chinese Acad Sci Inst Software Beijing Peoples R China Tsinghua Univ Dept Comp Sci & Technol Beijing 100084 Peoples R China Tsinghua Univ Minist Educ Key Lab Earth Syst Modeling Beijing 100084 Peoples R China Tsinghua Univ Ctr Earth Syst Sci Beijing 100084 Peoples R China Natl Univ Def Technol Dept Comp Sci & Technol Changsha Hunan Peoples R China Chinese Acad Sci State Key Lab Space Weather Beijing Peoples R China

ISBN: (纸本)9781450319225

Developing highly scalable algorithms for global atmospheric modeling is becoming increasingly important as scientists inquire to understand behaviors of the global atmosphere at extreme scales. Nowadays, heterogeneous architecture based on both processors and accelerators is becoming an important solution for large-scale computing. However, large-scale simulation of the global atmosphere brings a severe challenge to the development of highly scalable algorithms that fit well into state-of-the-art heterogeneous systems. Although successes have been made on GPU-accelerated computing in some top-level applications, studies on fully exploiting heterogeneous architectures in global atmospheric modeling are still very less to be seen, due in large part to both the computational difficulties of the mathematical models and the requirement of high accuracy for long term simulations. In this paper, we propose a peta-scalable hybrid algorithm that is successfully applied in a cubed-sphere shallow-water model for global atmospheric simulations. We employ an adjustable partition between CPUs and GPUs to achieve a balanced utilization of the entire hybrid system, and present a pipe-flow scheme to conduct conflict-free inter-node communication on the cubed-sphere geometry and to maximize communication-computation overlap. Systematic optimizations for multithreading on both GPU and CPU sides are performed to enhance computing throughput and improve memory efficiency. Our experiments demonstrate nearly ideal strong and weak scalabilities on up to 3,750 nodes of the Tianhe-1A. the largest run sustains a performance of 0.8 Pflops in double precision (32% of the peak performance), using 45,000 CPU cores and 3,750 GPUs.

关键词： parallel algorithm atmospheric modeling GPU heterogeneous system communication-computation overlap scalability

来源：评论

学校读者我要写书评

暂无评论

Gradient Compression Supercharged High-Performance Data parallel DNN Training 21

Gradient Compression Supercharged High-Performance Data Para...

引用

28th acm symposium on Operating Systems principles (SOSP)

作者： Bai, Youhui Li, Cheng Zhou, Quan Yi, Jun Gong, Ping Yan, Feng Chen, Ruichuan Xu, Yinlong Univ Sci & Technol China Hefei Peoples R China Univ Nevada Reno NV USA Nokia Bell Labs Murray Hill NJ USA Anhui Prov Key Lab High Performance Comp Hefei Peoples R China

ISBN: (纸本)9781450387095

Gradient compression is a promising approach to alleviating the communication bottleneck in data parallel deep neural network (DNN) training by significantly reducing the data volume of gradients for synchronization. While gradient compression is being actively adopted by the industry (e.g., Facebook and AWS), our study reveals that there are two critical but often overlooked challenges: 1) inefficient coordination between compression and communication during gradient synchronization incurs substantial overheads, and 2) developing, optimizing, and integrating gradient compression algorithms into DNN systems imposes heavy burdens on DNN practitioners, and ad-hoc compression implementations often yield surprisingly poor system performance. In this paper, we first propose a compression-aware gradient synchronization architecture, CaSync, which relies on a flexible composition of basic computing and communication primitives. It is general and compatible with any gradient compression algorithms and gradient synchronization strategies, and enables high-performance computation-communication pipelining. We further introduce a gradient compression toolkit, CompLL, to enable efficient development and automated integration of on-GPU compression algorithms into DNN systems with little programming burden. Lastly, we build a compression-aware DNN training framework HiPress with CaSync and CompLL. HiPress is open-sourced and runs on mainstream DNN systems such as MXNet, TensorFlow, and PyTorch. Evaluation via a 16-node cluster with 128 NVIDIA V100 GPUs and 100Gbps network shows that HiPress improves the training speed over current compression-enabled systems (e.g., BytePS-onebit and Ring-DGC) by 17.2%-69.5% across six popular DNN models.

关键词： DNN training gradient compression

来源：评论

学校读者我要写书评

暂无评论

Learning Modular programming using Realistic Scenarios: An Experience Report 21

Learning Modular Programming using Realistic Scenarios: An E...

引用

35th Brazilian symposium on Software Engineering, SBES 2021, held in conjunction with the Brazilian Conference on Software: theory and practice, CBSoft 2021

作者： Santos, Paulo Massollar, Jobson Federal University of the State of Rio de Janeiro Departament of Applied Informatics Rio de Janeiro Brazil

ISBN: (纸本)9781450390613

Software Engineering is essentially an applied research field. Students will only understand some of the abstract concepts and principles taught in software engineering courses if they see the intricacies of the real world in practice. Industrial or close-to-real scenarios are increasingly being incorporated into classrooms. It tends to enable the students to appreciate software development as an engineering process in which aspects such as quality must be considered in decision-making. this paper presents an instructional proposal adopted in the Modular programming course at Federal University of the State of Rio de Janeiro (UNIRIO). the proposed instructional design is based on Project-Based Learning theory, mixing group and individual tasks during the software development activities of a software project. It capitalizes on the microservice architectural pattern so that the students work in parallel developing different system modules. Also, to increase the perception of realism by the students, the instructional design employs modern software development tools commonly found in software organizations. Our preliminary results indicate students are more engaged in the course activities, and they feel more technically productive in their jobs. On the other hand, students pointed out some issues such as better accommodating task dependencies in the developed project and facilitating the configuration of software development tools. In conclusion, we believe that our proposal can be used in or adapted to other institutions interested in providing a more realistic experience to students in learning Modular programming. © 2021 acm.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Dynamic instrumentation of threaded applications

Proceedings of the ACM SIGPLAN Symposium on Principles and P...

引用

Proceedings of the acm SIGPLAN symposium on principles and practice of parallel programming, PPOPP 1999年 49-59页

作者： Xu, Zhichen Miller, Barton P. Naim, Oscar Univ of Wisconsin Madison WI United States

the design of non-trace based performance instrumentation techniques for threaded programs is investigated to provide detailed performance data while maintaining control of instrumentation costs. the design is based on low contention data structures. the Paradyn's dynamic instrumentation is extended to handle threaded programs. To associate data with individual threads, all threads must share the same instrumentation code and assign each thread with its own private copy of performance counters or timers. the asynchrony in a threaded program poses a major challenge to dynamic instrumentation.

关键词： parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

Evaluation of computing paradigms for n-body simulations on distributed memory architectures

Proceedings of the ACM SIGPLAN Symposium on Principles and P...

引用

Proceedings of the acm SIGPLAN symposium on principles and practice of parallel programming, PPOPP 1999年 25-36页

作者： McCurdy, Collin Mellor-Crummey, John Univ of Wisconsin Madison United States

the efficiency of HPF with respect to irregular applications is still largely unproven. While recent work has shown that a highly irregular hierarchical n-body force calculation method can be implemented in HPF, we have found that the implementation contains inefficiencies which cause it to run up to a factor of three times slower than our hand-coded, explicitly parallel implementation. Our work examines these inefficiencies, determines that most of the extra overhead is due to a single aspect of the communication strategy, and demonstrates that fixing the communication strategy can bring the overheads of the HPF application to within 25% of those of the hand-coded version.

关键词： parallel processing systems

来源：评论

学校读者我要写书评

暂无评论

the design and implementation of a parallel array operator for the arbitrary remapping of data

引用

acm SIGPLAN NOTICES 2003年第10期38卷 154-165页

作者： Deitz, SJ Chamberlain, BL Choi, SE Snyder, L Univ Washington Seattle WA 98195 USA Cray Inc Seattle WA 98104 USA Los Alamos Natl Lab Los Alamos NM 87545 USA

Gather and scatter are data redistribution functions of long-standing importance to high performance computing. In this paper, we present a highly-general array operator with powerful gather and scatter capabilities unmatched by other array languages. We discuss an efficient parallel implementation, introducing three new optimizations-schedule compression, dead array reuse, and direct communication-that reduce the costs associated with the operator's wide applicability. In our implementation of this operator in ZPL, we demonstrate performance comparable to the hand-coded Fortran + MPI versions of the NAS FT and CC benchmarks.

关键词： languages parallel programming gather scatter array languages ZPL

来源：评论

学校读者我要写书评

暂无评论

Efficient implementation scheme of concurrent object-oriented languages on stock multicomputers

Efficient implementation scheme of concurrent object-oriente...

引用

Proceedings of the 4th acm SIGPLAN symposium on principles & practice of parallel programming

作者： Taura, Kenjiro Matsuoka, Satoshi Yonezawa, Akinori Univ of Tokyo Japan

ISBN: (纸本)0897915895

Several novel techniques for efficient implementation of concurrent object-oriented languages on general purpose, stock multicomputers are presented. these techniques have been developed in implementing our concurrent object-oriented language ABCL on a Fujitsu Laboratory's experimental multicomputer AP1000 consisting of 512 SPARC chips. the proposed intra-node scheduling mechanism reduces the cost of local message passing. the cost of intra-node asynchronous message passing is about 20 SPARC instructions in the best case, including locality checking, dynamic method lookup, and scheduling. the minimum latency of asynchronous internode message passing is about 9 μs, or about 120 instructions, employing the self-dispatching mechanism independently proposed by Eicken et al. A large scale benchmark which involves 9,000,000 message passings shows 440 times speedup on the 512 nodes system compared to the sequential version of the same algorithm. We rely on simple hardware support for message passing and use no specialized architectural supports for object-oriented computing. thus, we are able to enjoy the benefits of future progress in standard processor technology. Our result shows that concurrent object-oriented languages can be implemented efficiently on conventional multicomputers.

关键词： Object oriented programming

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：