检索结果-内蒙古大学图书馆

COMPILING FOR distributed-memory systems

PROCEEDINGS OF THE IEEE 1993年第2期81卷 264-287页

作者： ZIMA, HP CHAPMAN, BM UNIV BONN W-5300 BONN GERMANY RICE UNIV HOUSTON TX 77251 USA

distributed-memory systems are potentially scalable to a very large number of processors and promise to be powerful tools for solving large-scale scientific and engineering problems. However, these machines are currently difficult to program, since the user has to distribute the data across the processors and explicitly formulate the communication required by the program under the selected distribution. During the past years, language extensions of standard programming languages such as Fortran were developed that allow a concise formulation of data distribution, and new compilation methods were designed and implemented that allow the programming of such machines at this relatively high level. In this paper, we describe the current state of the art in compiling procedural languages (in particular, Fortran) for distributed-memory machines, analyze the limitations of these approaches, and outline future research.

关键词： Data Parallel FORTRAN FORTRAN Message-Passing FORTRAN distributed-memory systems equivalent parallel message-passing programs extended FORTRAN 77 global addresses machine-independent language extension optimization methods parallel programming program compilers run-time analysis source-to-source translation

来源：评论

学校读者我要写书评

暂无评论

Balancing the performance of block multithreaded distributed-memory systems

引用

SIMULATION MODELLING PRACTICE AND THEORY 2011年第5期19卷 1318-1329页

作者： Zuberek, W. M. Memorial Univ Dept Comp Sci St John NF A1B 3X5 Canada Univ Life Sci Dept Appl Informat Warsaw Poland

The performance of modern computer systems is increasingly often limited by long latencies of accesses to the memory subsystems. Instruction-level multithreading is an architectural approach to tolerating such long latencies by switching instruction threads rather than waiting for the completion of memory operations. The paper studies performance limitations in distributed-memory block multithreaded systems and determines conditions for such systems to be balanced. Event-driven simulation of a timed Petri net model of a simple distributed-memory system confirms the derived performance results. (C) 2011 Elsevier B.V. All rights reserved.

关键词： Block multithreading distributed-memory systems Performance analysis Balanced systems Timed Petri nets

来源：评论

学校读者我要写书评

暂无评论

A general approach for supporting nonblocking data structures on distributed-memory systems

引用

JOURNAL OF PARALLEL AND distributed COMPUTING 2023年 173卷 48-60页

作者： Thanh-Dang Diep Phuong Hoai Ha Fuerlinger, Karl Ludwig Maximilians Univ LMU Munich Comp Sci Dept MNM Team Oettingenstr 67 D-80538 Munich Germany UiT Arctic Univ Norway Dept Comp Sci NO-9037 Tromso Norway

Nonblocking data structures are an essential part of many parallel applications in that they can help to improve fault tolerance and performance. Although there are scores of nonblocking data structures, such as stacks, queues, double-ended queues (deques), lists, widely used in practice, most of them are designed to be used on shared-memory machines only, and cannot be used in a distributed-memory setting. Several recent studies focus on the development of novel tailor-made nonblocking distributed data structures and omit the potential for adapting a great wealth of existing nonblocking shared -memory ones for the distributed-memory case. Hence, we propose a general approach for bridging the gap between most existing nonblocking data structures and distributed-memory machines in this work. Several challenges, such as safe memory reclamation and solving the ABA problem, must be overcome. To address these issues, we present a global memory management scheme. The scheme takes advantage of hazard pointers which are widely used to tackle the problems in shared-memory environments. To demonstrate our general approach, we take stacks as a typical example of nonblocking data structures. This work also provides a survey of well-known nonblocking stack algorithms along with our analysis and evaluation in distributed-memory environments. Moreover, this paper depicts how to improve performance of a stack algorithm by making use of node locality. Besides, a cost model based on worst cases is devised to help gain a better understanding into experimental results of nonblocking distributed data structures, along with our analysis of the influence of two popular lock-free programming patterns on performance. (c) 2022 Elsevier Inc. All rights reserved.

关键词： Nonblocking data structures distributed-memory systems Nonblocking stacks Global memory reclamation ABA problem

来源：评论

学校读者我要写书评

暂无评论

A high-speed network interface for distributed-memory systems: Architecture and applications

引用

ACM TRANSACTIONS ON COMPUTER systems 1997年第1期15卷 75-109页

作者： Steenkiste, P Carnegie Mellon Univ. Pittsburgh PA

distributed-memory systems have traditionally had great difficulty performing network I/O at rates proportional to their computational power. The problem is that the network interface has to support network I/O for a supercomputer, using computational and memory bandwidth resources similar to those of a workstation. As a result, the network interface becomes a bottleneck. In this article we present an I/O architecture that addresses these problems and supports high-speed network I/O on distributed-memory systems. The key to good performance is to partition the work appropriately between the system and the network interface. Some communication tasks are performed on the distributed-memory parallel system, since it is more powerful and less likely to become a bottleneck than the network interface. Tasks that do not parallelize well are performed on the network interface, and hardware support is provided for the most time-critical operations. This architecture has been implemented for the iWarp distributed-memory system and has been used by a number of applications. We describe this implementation, present performance results, and use application examples to validate the main features of the I/O architecture.

关键词： application-managed I/O data reshuffling distributed-memory systems I/O architecture network interface outboard buffering protocol processing resource management

来源：评论

学校读者我要写书评

暂无评论

On-line error detection through data duplication in distributed-memory systems

引用

MICROPROCESSORS AND MICROsystems 1997年第3期21卷 197-209页

作者： Gong, C Melhem, R Gupta, R Univ Pittsburgh Dept Comp Sci Pittsburgh PA 15260 USA

The scalability of distributed-memory systems makes them attractive for massively parallel computations, However, the issues of fault detection and fault tolerance are critical in such systems since the probability of having faulty components increases with the number of processors. We propose a methodology for fault detection on distributed-memory systems through compiler support. The single-program multiple-data (SPMD) execution model is extended to execute programs in which selected data items are duplicated on different processors, During execution, whenever the values of duplicated data are computed, they are compared for the purpose of error detection. In other words, fault detection is controlled by the duplication of data. The proposed compiler-assisted fault-detection technique does not require any specialized hardware and attempts to exploit the idle capacity of the system whenever possible. After presenting the principles of duplicated computation through data duplication and the corresponding compiler algorithms, we focus on regular loops to exploit the idle processors for fault-detection purpose. We present experimental results that demonstrate the feasibility of the proposed approach. (C) 1997 Elsevier Science B.V.

关键词： parallel computations distributed-memory systems processor

来源：评论

学校读者我要写书评

暂无评论

Parallel H-matrix arithmetic on distributed-memory systems

引用

COMPUTING AND VISUALIZATION IN SCIENCE 2012年第2期15卷 87-97页

作者： Izadi, Mohammad Shahid Bahonar Univ Kerman Dept Math POB 76169-14111 Kerman Iran

In the last decade, the hierarchical matrix technique was introduced to deal with dense matrices in an efficient way. It provides a data-sparse format and allows an approximate matrix algebra of nearly optimal complexity. This paper is concerned with utilizing multiple processors to gain further speedup for the H-matrix algebra, namely matrix truncation, matrix-vector multiplication, matrix-matrix multiplication, and inversion. One of the most cost-effective solution for large-scale computation is distributed computing. Distribute-memory architectures provide an inexpensive way for an organization to obtain parallel capabilities as they are increasingly popular. In this paper, we introduce a new distribution scheme for H-matrices based on the corresponding index set. Numerical experiments applied to a BEM model will complement our complexity analysis.

关键词： Hierarchical matrices Parallel algorithm distributed-memory systems

来源：评论

学校读者我要写书评

暂无评论

Massively Parallel Polar Decomposition on distributed-memory systems

引用

ACM TRANSACTIONS ON PARALLEL COMPUTING 2019年第1期6卷 1–15页

作者： Ltaief, Hatem Sukkari, Dalal Esposito, Aniello Nakatsukasa, Yuji Keyes, David King Abdullah Univ Sci & Technol Extreme Comp Res Ctr 4700 King Abdullah Blvd Jeddah 23955 Saudi Arabia Cray EMEA Res Lab Bristol Avon England Univ Oxford Math Inst Oxford England

We present a high-performance implementation of the Polar Decomposition (PD) on distributed-memory systems. Building upon on the QR-based Dynamically Weighted Halley (QDWH) algorithm, the key idea lies in finding the best rational approximation for the scalar sign function, which also corresponds to the polar factor for symmetric matrices, to further accelerate the QDWH convergence. Based on the Zolotarev rational functions-introduced by Zolotarev (ZOLO) in 1877-this new PD algorithm ZOLO-PD converges within two iterations even for ill-conditioned matrices, instead of the original six iterations needed for QDWH. ZOLO-PD uses the property of Zolotarev functions that optimality is maintained when two functions are composed in an appropriate manner. The resulting ZOLO-PD has a convergence rate up to 17, in contrast to the cubic convergence rate for QDWH. This comes at the price of higher arithmetic costs and memory footprint. These extra floating-point operations can, however, be processed in an embarrassingly parallel fashion. We demonstrate performance using up to 102,400 cores on two supercomputers. We demonstrate that, in the presence of a large number of processing units, ZOLO-PD is able to outperform QDWH by up to 2.3x speedup, especially in situations where QDWH runs out of work, for instance, in the strong scaling mode of operation.

关键词： Polar decomposition Zolotarev functions parallel algorithms strong scaling distributed-memory systems

来源：评论

学校读者我要写书评

暂无评论

Scalable Triangle Counting on distributed-memory systems

Scalable Triangle Counting on Distributed-Memory Systems

引用

IEEE High Performance Extreme Computing Conference (HPEC)

作者： Acer, Seher Yasar, Abdurrahman Rajamanickam, Sivasankaran Wolf, Michael Catalyurek, Umit V. Sandia Natl Labs Ctr Comp Res POB 5800 Albuquerque NM 87185 USA Georgia Inst Technol Coll Comp Atlanta GA 30332 USA

ISBN: (纸本)9781728150208

Triangle counting is a foundational graph-analysis kernel in network science. It has also been one of the challenge problems for the "Static Graph Challenge". In this work, we propose a novel, hybrid, parallel triangle counting algorithm based on its linear algebra formulation. Our framework uses MPI and Cilk to exploit the benefits of distributed-memory and shared-memory parallelism, respectively. The problem is partitioned among MPI processes using a two-dimensional (2D) Cartesian block partitioning. One-dimensional (1D) row-wise partitioning is used within the Cartesian blocks for shared-memory parallelism using the Cilk programming model. Besides exhibiting very good strong scaling behavior in almost all tested graphs, our algorithm achieves the fastest time on the 1.4B edge real-world twitter graph, which is 3.217 seconds, on 1,092 cores. In comparison to past distributed-memory parallel winners of the graph challenge, we demonstrate a speed up of 2.7x on this twitter graph. This is also the fastest time reported for parallel triangle counting on the twitter graph when the graph is not replicated.

关键词： triangle counting distributed-memory systems scale-free graphs two-dimensional partitioning

来源：评论

学校读者我要写书评

暂无评论

Sphynx: A parallel multi-GPU graph partitioner for distributed-memory systems

引用

PARALLEL COMPUTING 2021年 106卷 102769-102769页

作者： Acer, Seher Boman, Erik G. Glusa, Christian A. Rajamanickam, Sivasankaran Sandia Natl Labs Ctr Comp Res POB 5800 Albuquerque NM 87185 USA

Graph partitioning has been an important tool to partition the work among several processors to minimize the communication cost and balance the workload. While accelerator-based supercomputers are emerging to be the standard, the use of graph partitioning becomes even more important as applications are rapidly moving to these architectures. However, there is no distributed-memory-parallel, multi-GPU graph partitioner available for applications. We developed a spectral graph partitioner, Sphynx, using the portable, accelerator friendly stack of the Trilinos framework. In Sphynx, we allow using different preconditioners and exploit their unique advantages. We use Sphynx to systematically evaluate the various algorithmic choices in spectral partitioning with a focus on the GPU performance. We perform those evaluations on two distinct classes of graphs: regular (such as meshes, matrices from finite element methods) and irregular (such as social networks and web graphs), and show that different settings and preconditioners are needed for these graph classes. The experimental results on the Summit supercomputer show that Sphynx is the fastest alternative on irregular graphs in an application-friendly setting and obtains a partitioning quality close to ParMETIS on regular graphs. When compared to nvGRAPH on a single GPU, Sphynx is faster and obtains better balance and better quality partitions. Sphynx provides a good and robust partitioning method across a wide range of graphs for applications looking for a GPU-based partitioner.

关键词： Graph partitioning Spectral partitioning GPUs distributed-memory systems

来源：评论

学校读者我要写书评

暂无评论

Stochastic Gradient Descent for matrix completion: Hybrid parallelization on shared- and distributed-memory systems

引用

KNOWLEDGE-BASED systems 2024年 283卷

作者： Buyukkaya, Kemal Karsavuran, M. Ozan Aykanat, Cevdet Bilkent Univ Comp Engn Dept Turkiye TR-06800 Ankara Turkiye Lawrence Berkeley Natl Lab 1 Cyclotron Rd Berkeley CA 94720 USA

The purpose of this study is to investigate the hybrid parallelization of the Stochastic Gradient Descent (SGD) algorithm for solving the matrix completion problem on a high-performance computing platform. We propose a hybrid parallel decentralized SGD framework with asynchronous inter-process communication and a novel flexible partitioning scheme to attain scalability up to hundreds of processors. We utilize Message Passing Interface (MPI) for inter-node communication and POSIX threads for intra-node parallelism. We tested our method by using different real-world benchmark datasets. Experimental results on a hybrid parallel architecture showed that, compared to the state-of-the-art, the proposed algorithm achieves 6x higher throughput on sparse datasets, while it achieves comparable throughput on relatively dense datasets.

关键词： Stochastic gradient descent Matrix completion Collaborative filtering Matrix factorization distributed-memory systems Shared-memory systems Hybrid parallelism

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：