检索结果-内蒙古大学图书馆

Optimizing memory transactions for large-scale programs

JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING 2016年 89卷 13-24页

作者： Carvalho, Fernando Miguel Cachopo, Joao Univ Lisbon INESC ID Inst Super Tecn P-1699 Lisbon Portugal Polytech Inst Lisbon Inst Super Engn Lisboa ADEETC Lisbon Portugal

Even though Software Transactional Memory (STM) is one of the most promising approaches to simplify concurrent programming, current STM implementations incur significant overheads that render them impractical for many real-sized programs. The key insight of this work is that we do not need to use the same costly barriers for all the memory managed by a real-sized application, if only a small fraction of the memory is under contention lightweight barriers may be used in this case. In this work, we propose a new solution based on an approach of adaptive object metadata (AOM) to promote the use of a fast path to access objects that are not under contention. We show that this approach is able to make the performance of an STM competitive with the best fine-grained lock-based approaches in some of the more challenging benchmarks. (C) 2015 Elsevier Inc. All rights reserved.

关键词： Software Transactional Memory Runtime optimizations concurrent programming

来源：评论

学校读者我要写书评

暂无评论

MPI-ACC: Accelerator-Aware MPI for Scientific Applications

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2016年第5期27卷 1401-1414页

作者： Aji, Ashwin M. Panwar, Lokendra S. Ji, Feng Murthy, Karthik Chabbi, Milind Balaji, Pavan Bisset, Keith R. Dinan, James Feng, Wu-chun Mellor-Crummey, John Ma, Xiaosong Thakur, Rajeev Virginia Tech Dept Comp Sci Blacksburg VA 24061 USA N Carolina State Univ Dept Comp Sci Raleigh NC 27695 USA Rice Univ Dept Comp Sci Houston TX USA Argonne Natl Lab Math & Comp Sci Argonne IL 60439 USA Virginia Tech Virginia Bioinformat Inst Blacksburg VA USA

Data movement in high-performance computing systems accelerated by graphics processing units (GPUs) remains a challenging problem. Data communication in popular parallel programming models, such as the Message Passing Interface (MPI), is currently limited to the data stored in the CPU memory space. Auxiliary memory systems, such as GPU memory, are not integrated into such data movement standards, thus providing applications with no direct mechanism to perform end-to-end data movement. We introduce MPI-ACC, an integrated and extensible framework that allows end-to-end data movement in accelerator-based systems. MPI-ACC provides productivity and performance benefits by integrating support for auxiliary memory spaces into MPI. MPI-ACC supports data transfer among CUDA, OpenCL and CPU memory spaces and is extensible to other offload models as well. MPI-ACC's runtime system enables several key optimizations, including pipelining of data transfers, scalable memory management techniques, and balancing of communication based on accelerator and node architecture. MPI-ACC is designed to work concurrently with other GPU workloads with minimum contention. We describe how MPI-ACC can be used to design new communication-computation patterns in scientific applications from domains such as epidemiology simulation and seismology modeling, and we discuss the lessons learned. We present experimental results on a state-of-the-art cluster with hundreds of GPUs;and we compare the performance and productivity of MPI-ACC with MVAPICH, a popular CUDA-aware MPI solution. MPI-ACC encourages programmers to explore novel application-specific optimizations for improved overall cluster utilization.

关键词： Heterogeneous (hybrid) systems parallel systems distributed architectures concurrent programming

来源：评论

学校读者我要写书评

暂无评论

Universal constructions that ensure disjoint-access parallelism and wait-freedom

引用

DISTRIBUTED COMPUTING 2016年第4期29卷 251-277页

作者： Ellen, Faith Fatourou, Panagiota Kosmas, Eleftherios Milani, Alessia Travers, Corentin Univ Toronto 10 Kings Coll Rd Toronto ON Canada Fdn Res & Technol Hellas FORTH Inst Comp Sci N Plastira 100 Iraklion 70013 Crete Island Greece Univ Crete Voutes Campus Iraklion 70013 Crete Island Greece Univ Bordeaux LaBRI F-33400 Talence France CNRS F-33400 Talence France

A universal construction is a general mechanism for obtaining a concurrent implementation of an object from its sequential code. We show that there is no universal construction that is both disjoint-access parallel (guaranteeing the processes operating on different parts of an implemented object do not interfere with one another) and wait-free (guaranteeing progress for each nonfaulty process when accessing an object). In contrast, we present a universal construction which results in disjoint-access parallel, wait-free implementations of any object provided there is a bound on the number of data items accessed by each operation supported by the object.

关键词： concurrent programming Disjoint-access parallelism Wait-freedom Universal construction Impossibility result

来源：评论

学校读者我要写书评

暂无评论

Highly scalable Erlang framework for agent-based metaheuristic computing

引用

JOURNAL OF COMPUTATIONAL SCIENCE 2016年第Part1期17卷 234-248页

作者： Turek, Wojciech Stypka, Jan Krzywicki, Daniel Anielski, Piotr Pietak, Kamil Byrski, Aleksander Kisiel-Dorohinicki, Marek AGH Univ Sci & Technol Fac Comp Sci Elect & Telecommun Al Mickiewicza 30 PL-30059 Krakow Poland

Difficult search and optimization problems, usually solved by metaheuristics, are very often implemented in concurrent and parallel environment, as many metaheuristics (e.g. population- or agent-based) are inherently easy to parallelize. Therefore search for easy-to-use, robust and efficient frameworks dedicated for such computing methods, especially in the era of ubiquitous many and multi-core systems, is very desirable. Indeed, the development of multi-core architectures is incredibly fast and multicore CPUs can be found nowadays not only in supercomputers, but also in ordinary laptops or even phones. Efficient use of multicore architectures requires applying suitable languages and technologies, like Erlang. Its concurrency model, based on lightweight processes and asynchronous message-passing, seems very well suited for running massively concurrent code on many cores. Most of existing Erlang industrial applications are deployed on computers with up to 24 CPU cores, and there are hardly any reports on using Erlang on architectures exceeding 32 physical cores. In this paper we present our experiences with developing a concurrent Erlang-based computing platform, scaling computationally-intensive and memory-intensive applications up to 64 cores, using as examples global optimization and urban traffic planning problems. (C) 2016 Elsevier B.V. All rights reserved.

关键词： Metaheuristic computing concurrent programming Scalability Erlang

来源：评论

学校读者我要写书评

暂无评论

Generalised rely-guarantee concurrency: an algebraic foundation

引用

FORMAL ASPECTS OF COMPUTING 2016年第6期28卷 1057-1078页

作者： Hayes, Ian J. Univ Queensland Sch Informat Technol & Elect Engn Brisbane Qld Australia

The rely-guarantee technique allows one to reason compositionally about concurrent programs. To handle interference the technique makes use of rely and guarantee conditions, both of which are binary relations on states. A rely condition is an assumption that the environment performs only atomic steps satisfying the rely relation and a guarantee is a commitment that every atomic step the program makes satisfies the guarantee relation. In order to investigate rely-guarantee reasoning more generally, in this paper we allow interference to be represented by a process rather than a relation and hence derive more general rely-guarantee laws. The paper makes use of a weak conjunction operator between processes, which generalises a guarantee relation to a guarantee process, and introduces a rely quotient operator, which generalises a rely relation to a process. The paper focuses on the algebraic properties of the general rely-guarantee theory. The Jones-style rely-guarantee theory can be interpreted as a model of the general algebraic theory and hence the general laws presented here hold for that theory.

关键词： concurrent programming rely-guarantee concurrency program verification program algebra concurrent Kleene algebra

来源：评论

学校读者我要写书评

暂无评论

PathGraph: A Path Centric Graph Processing System

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2016年第10期27卷 2998-3012页

作者： Yuan, Pingpeng Xie, Changfeng Liu, Ling Jin, Hai Huazhong Univ Sci & Technol Serv Comp Technol & Syst Lab Cluster & Grid Comp Lab Big Data Technol & Syst LabSch Comp Sci & Techno Wuhan 430074 Peoples R China Georgia Inst Technol Coll Comp Atlanta GA 30332 USA

Large scale iterative graph computation presents an interesting systems challenge due to two well known problems: (1) the lack of access locality and (2) the lack of storage efficiency. This paper presents PathGraph, a system for improving iterative graph computation on graphs with billions of edges. First, we improve the memory and disk access locality for iterative computation algorithms on large graphs by modeling a large graph using a collection of tree-based partitions. This enables us to use path-centric computation rather than vertex-centric or edge-centric computation. For each tree partition, we re-label vertices using DFS in order to preserve consistency between the order of vertex ids and vertex order in the paths. Second, a compact storage that is optimized for iterative graph parallel computation is developed in the PathGraph system. Concretely, we employ delta-compression and store tree-based partitions in a DFS order. By clustering highly correlated paths together as tree based partitions, we maximize sequential access and minimize random access on storage media. Third but not the least, our path-centric computation model is implemented using a scatter/gather programming model. We parallel the iterative computation at partition tree level and perform sequential local updates for vertices in each tree partition to improve the convergence speed. To provide well balanced workloads among parallel threads at tree partition level, we introduce the concept of multiple stealing points based task queue to allow work stealings from multiple points in the task queue. We evaluate the effectiveness of PathGraph by comparing with recent representative graph processing systems such as GraphChi and X-Stream etc. Our experimental results show that our approach outperforms the two systems on a number of graph algorithms for both in-memory and out-of-core graphs. While our approach achieves better data balance and load balance, it also shows better speedup than the two

关键词： Graphs and networks concurrent programming graph algorithms data storage representations

来源：评论

学校读者我要写书评

暂无评论

Network Traffic Processing With PFQ

引用

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS 2016年第6期34卷 1819-1833页

作者： Bonelli, Nicola Giordano, Stefano Procissi, Gregorio Univ Pisa Dipartimento Ingn Informaz I-56122 Pisa Italy Consorzio Nazl Interuniv Telecomunicaz I-56122 Pisa Italy

This paper presents Packet Family Queue (PFQ), a high-performance framework for packet processing designed to flexibly handle network applications parallelism and making traffic processing safe and easy. PFQ is an open-source module for the Linux kernel that combines software-accelerated packet I/O to in-kernel early stage packet processing and fine-grained distribution to network applications and physical devices. PFQ does not require any modification to network device drivers and exposes programming interfaces to multi-threaded applications natively designed to run on top of it, as well as to legacy monitoring tools using the pcap library. The results show that the flexibility and the backward compatibility provided by PFQ do not impact its processing performance that, in fact, reaches line rate figures in the cases of pure speed tests and real practical monitoring use cases on 10+Gb/s links.

关键词： Network monitoring concurrent programming multi-core architectures multi-queue NICs early stage processing application offload

来源：评论

学校读者我要写书评

暂无评论

Automated Data Partitioning for Highly Scalable and Strongly Consistent Transactions

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2016年第1期27卷 106-118页

作者： Turcu, Alexandru Palmieri, Roberto Ravindran, Binoy Hirve, Sachin Virginia Tech Dept Elect & Comp Engn Blacksburg VA 24060 USA

Modern transactional processing systems need to be fast and scalable, but this means many such systems settled for weak consistency models. It is however possible to achieve all of strong consistency, high scalability and high performance, by using fine-grained partitions and light-weight concurrency control that avoids superfluous synchronization and other overheads such as lock management. Independent transactions are one such mechanism, that rely on good partitions and appropriately defined transactions. On the downside, it is not usually straightforward to determine optimal partitioning schemes, especially when dealing with non-trivial amounts of data. Our work attempts to solve this problem by automating the partitioning process, choosing the correct transactional primitive, and routing transactions appropriately.

关键词： Distributed databases code generation concurrent programming

来源：评论

学校读者我要写书评

暂无评论

WFR-TM: Wait-free readers without sacrificing speculation of writers

引用

JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING 2016年 96卷 134-151页

作者： Fatourou, Panagiota Kanellou, Eleni Kosmas, Eleftherios Rabbi, Md Forhad Fdn Res & Technol Hellas FORTH ICS N Plastira 100 GR-70013 Iraklion Crete Island Greece Univ Crete Dept Comp Sci POB 2208Voutes Campus GR-70013 Iraklion Crete Island Greece Curtin Univ Dept Comp Bentley WA Australia

Transactional Memory (TM) is a promising concurrent programming paradigm which employs transactions to achieve synchronization in accessing common data known as transactional variables. A transaction may either commit, making its updates to transactional variables visible, or abort, discarding its updates. We introduce WFR-TM, a TM algorithm which attempts to combine the advantages of pessimistic and optimistic TM. In a pessimistic TM, no transaction ever aborts;however, current pessimistic TM implementations, execute update transactions sequentially, decreasing the degree of achieved parallelism. In optimistic TM, transactions are executed concurrently and they commit if they have encountered no conflict during their execution. In WFR-TM, read-only transactions not only are wait-free, but also they never execute expensive synchronization operations (like CAS, LL/SC, etc.). This is achieved without sacrificing the parallelism between update transactions. Update transactions synchronize pessimistically with concurrently executed read-only transactions but they synchronize optimistically with each other. (C) 2016 Elsevier Inc.

关键词： concurrent programming Transactional memory Wait-freedom Optimistic TM Pessimistic TM

来源：评论

学校读者我要写书评

暂无评论

Dynamic slicing of concurrent specification languages

引用

PARALLEL COMPUTING 2016年第Apr.期53卷 1-22页

作者： Llorens, M. Oliver, J. Silva, J. Tamarit, S. Univ Politecn Valencia Dept Sistemas Informat & Comp E-46022 Valencia Spain Univ Politecn Madrid Babel Res Grp Fac Informat Campus Montegancedo S-N Boadilla Del Monte 28660 Spain

Dynamic slicing is a technique to extract the part of the program (called slice) that influences or is influenced, in a particular execution, by a given point of interest in the source code (called slicing criterion). Since a single execution is considered, the technique often uses a trace of this execution to analyze data and control dependencies. In this work we present the first formulation and implementation of dynamic slicing in the context cif CSP. Most of the ideas presented can be directly applied to other concurrent specification languages such as Promela or CCS, but we center the discussion and the implementation on CSP. We base our technique on a new data structure to represent CSP computations called track. A track is a data structure which represents the sequence of expressions that have been evaluated during the computation, and moreover, it is labeled with the location of these expressions in the specification. The implementation of a dynamic slicer for CSP is useful for debugging, program comprehension, and program specialization, and it is also interesting from a theoretical perspective because CSP introduces difficulties such as heavy concurrency and non-determinism, synchronizations, frequent absence of data dependence, etc. (C) 2016 Elsevier B.V. All rights reserved.

关键词： concurrent programming CSP Slicing

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：