检索结果-内蒙古大学图书馆

A Compiler for Throughput Optimization of Graph Algorithms on GPUs

ACM SIGPLAN NOTICES 2016年第10期51卷 1-19页

作者： Pai, Sreepathi Pingali, Keshav Univ Texas Austin Austin TX 78712 USA

Writing high-performance GPU implementations of graph algorithms can be challenging. In this paper, we argue that three optimizations called throughput optimizations are key to high-performance for this application class. These optimizations describe a large implementation space making it unrealistic for programmers to implement them by hand. To address this problem, we have implemented these optimizations in a compiler that produces CUDA code from an intermediate-level program representation called IrGL. Compared to state-of-the-art handwritten CUDA implementations of eight graph applications, code generated by the IrGL compiler is up to 5.95x times faster (median 1.4x) for five applications and never more than 30% slower for the others. Throughput optimizations contribute an improvement up to 4.16x (median 1.4x) to the performance of unoptimized IrGL code.

关键词： Graph applications amorphous data-parallelism GPUs compilers optimization throughput

来源：评论

学校读者我要写书评

暂无评论

Execution of structured analysis specifications with an object oriented Petri net approach

Execution of structured analysis specifications with an obje...

引用

International conference on Computer languages

作者： P.J. Pulli Computer Technology Laboratory Technical Research Centre of Finland (VTT) Oulu Finland Institute of Electronics System Techniques Group Swiss Federal Institute of Technology Zurich Switzerland

The dynamic behavior of the basic modeling elements of the real-time structured-analysis specification method has been defined with an object-oriented approach as a hierarchy of classes of Smalltalk-inscribed Petri nets. As a result of this definition, a given instance of a structured-analysis basic modeling element can be derived as an instance of the corresponding net class. These instances can be connected together to form an equivalent net to a given structured-analysis specification. A scheduling algorithm has been defined for Smalltalk-inscribed Petri nets meeting the execution rules for real-time structured analysis. This algorithm has also been implemented in the SPECS specification and prototyping environment, thus allowing the graphic execution of structured-analysis specifications.< >

关键词： object oriented modeling Petri nets Real time systems Prototypes Scheduling algorithm Algorithm design and analysis Graphics object oriented programming Data processing Automata

来源：评论

学校读者我要写书评

暂无评论

A model for development of customizable applications

A model for development of customizable applications

引用

IEEE Annual International Computer Software and applications conference (COMPSAC)

作者： H.M. Atchan R. Bell Corporate Systems Development Division Honeywell Inc. Golden Valley MN USA

A model is presented that recognizes a multilevel development paradigm with participation of system developers (vendor), application specialists (vendor or customer), and end users (customer). Within this paradigm, the authors utilize several reusability technologies but adapt them to a range of developer skill levels. This model can be used for tools that allow reusability while facilitating customization of software applications and software products. It is also shown how the Application Development Toolkit and its underlying application development model address these issues to provide a real solution to the customization problem.< >

关键词： Application software Computer languages programming profession Database systems Libraries object oriented programming High level languages Software reusability Production systems Data mining

来源：评论

学校读者我要写书评

暂无评论

An execution model for distributed object-oriented computation 88

An execution model for distributed object-oriented computati...

引用

conference proceedings on object-oriented programming systems, languages and applications

作者： Edward H. Bensley Thomas J. Brando Myra Jean Prelle The MITRE Corporation Bedford MA

ISBN: (纸本)9780897912846

This paper describes an execution model being developed for distributed object-oriented in a message-passing multiple-instruction/multiple-data-stream (MIMD) environment. The objective is to execute an object-oriented program as concurrently as possible. Some opportunities for concurrency can be identified explicitly by the programmer. Others can be identified at compile time. There are some opportunities for concurrency, however, that can only be discovered at runtime because they are data-dependent. The model of execution we describe attempts to discover and exploit the data-dependent concurrency that exists in a given program execution.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Scalable Race Detection for Android applications

引用

ACM SIGPLAN NOTICES 2015年第10期50卷 332-348页

作者： Bielik, Pavol Raychev, Veselin Vechev, Martin Swiss Fed Inst Technol Dept Comp Sci Zurich Switzerland

We present a complete end-to-end dynamic analysis system for finding data races in mobile Android applications. The capabilities of our system significantly exceed the state of the art: our system can analyze real-world application interactions in minutes rather than hours, finds errors inherently beyond the reach of existing approaches, while still (critically) reporting very few false positives. Our system is based on three key concepts: (i) a thorough happens-before model of Android-specific concurrency, (ii) a scalable analysis algorithm for efficiently building and querying the happens-before graph, and (iii) an effective set of domain-specific filters that reduce the number of reported data races by several orders of magnitude. We evaluated the usability and performance of our system on 3 5 4 real-world Android applications (e.g., Facebook). Our system analyzes a minute of end-user interaction with the application in about 2 4 seconds, while current approaches take hours to complete. Inspecting the results for 8 large open-source applications revealed 1 5 harmful bugs of diverse kinds. Some of the bugs we reported were confirmed and fixed by developers.

关键词： Data Races Happens-before Android Non-determinism

来源：评论

学校读者我要写书评

暂无评论

Fast Splittable Pseudorandom Number Generators

引用

ACM SIGPLAN NOTICES 2014年第10期49卷 453-472页

作者： Steele, Guy L., Jr. Lea, Doug Flood, Christine H. SUNY Coll Oswego Oswego NY USA Red Hat Inc Raleigh NC USA

We describe a new algorithm SPLITMIX for an object-oriented and splittable pseudorandom number generator (PRNG) that is quite fast: 9 64-bit arithmetic/logical operations per 64 bits generated. A conventional linear PRNG object provides a generate method that returns one pseudorandom value and updates the state of the PRNG, but a splittable PRNG object also has a second operation, split, that replaces the original PRNG object with two (seemingly) independent PRNG objects, by creating and returning a new such object and updating the state of the original object. Splittable PRNG objects make it easy to organize the use of pseudorandom numbers in multithreaded programs structured using fork-join parallelism. No locking or synchronization is required (other than the usual memory fence immediately after object creation). Because the generate method has no loops or conditionals, it is suitable for SIMD or GPU implementation. We derive SPLITMIX from the DOTMIX algorithm of Leiserson, Schardl, and Sukha by making a series of program transformations and engineering improvements. The end result is an object-oriented version of the purely functional API used in the Haskell library for over a decade, but SPLITMIX is faster and produces pseudorandom sequences of higher quality;it is also far superior in quality and speed to java util. Random, and has been included in Java JDK8 as the class j ava. util. SplittableRandom. We have tested the pseudorandom sequences produced by SPLITMIX using two standard statistical test suites (DieHarder and TestU01) and they appear to be adequate for "everyday" use, such as in Monte Carlo algorithms and randomized data structures where speed is important.

关键词： Algorithms Performance collections determinism Java multithreading nondeterminism object-oriented parallel computing pedigree pseudorandom random number generator recursive splitting Scala spliterator splittable data structures streams

来源：评论

学校读者我要写书评

暂无评论

Gentrification Gone too Far? Affordable 2nd-Class Values for Fun and (Co-)Effect

引用

ACM SIGPLAN NOTICES 2016年第10期51卷 234-251页

作者： Osvald, Leo Essertel, Gregory Wu, Xilun Gonzalez-Alayon, Lilliam I. Rompf, Tiark Purdue Univ W Lafayette IN 47907 USA

First-class functions dramatically increase expressiveness, at the expense of static guarantees. In ALGOL or PASCAL, functions could be passed as arguments but never escape their defining scope. Therefore, function arguments could serve as temporary access tokens or capabilities, enabling callees to perform some action, but only for the duration of the call. In modern languages, such programming patterns are no longer available. The central thrust of this paper is to re-introduce second-class functions and other values alongside first-class entities in modern languages. We formalize second-class values with stack-bounded lifetimes as an extension to simply-typed lambda calculus, and for richer type systems such as F-<: and systems with path-dependent types. We generalize the binary first-vs second-class distinction to arbitrary privilege lattices, with the underlying type lattice as a special case. In this setting, abstract types naturally enable privilege parametricity. We prove type soundness and lifetime properties in Coq. We implement our system as an extension of Scala, and present several case studies. First, we modify the Scala Collections library and add privilege annotations to all higher-order functions. Privilege parametricity is key to retain the high degree of code-reuse between sequential and parallel as well as lazy and eager collections. Second, we use scoped capabilities to introduce a model of checked exceptions in the Scala library, with only few changes to the code. Third, we employ second-class capabilities for memory safety in a region-based off-heap memory library.

关键词： first-class second-class types effects capabilities object lifetimes

来源：评论

学校读者我要写书评

暂无评论

Coroutine sequencing in BETA

Coroutine sequencing in BETA

引用

Annual Hawaii International conference on System Sciences (HICSS)

作者： B.B. Kristensen O.L. Madsen B. Moller-Pedersen K. Nygaard Institute of Electronic Systems Aalborg University Center Computer Science Department Aarhus University Norsk Regnesentral-Norwegian Computing Center Norwegian Computing Center Oslo University

In object-oriented programming, a program execution is viewed as a physical model of some real or imaginary part of the world. A language supporting object-oriented programming must therefore contain comprehensive facilities for modeling phenomena and concepts form the application domain. Many applications in the real world consist of objects carrying out sequential processes. Coroutines may be used for modeling objects that alternate between a number of sequential processes. The authors describe coroutines in BETA.< >

关键词： object oriented modeling Computer languages object oriented programming Physics computing Computer science Informatics Modular construction Pattern matching

来源：评论

学校读者我要写书评

暂无评论

Shared memory multiprocessors and sequential programming languages: a case study

Shared memory multiprocessors and sequential programming lan...

引用

Annual Hawaii International conference on System Sciences (HICSS)

作者： L.A. Crowl Computer Science Department University of Rochester USA

The Chrysalis operating system for the BBN Butterfly parallel processor provides a flexible, object-oriented, shared-memory environment for parallel programming. C++ is a safe, sequential, object-oriented programming language. The pairing of these environments is seen potentially as a natural fit. The development of Chrysalis++, and interface between C++ and Chrysalis, indicates that subtle assumptions within each system lead to surprising inconsistencies. Any combination of sequential programming language and shared-memory multiprocessor operating system is likely to be plagued by these inconsistencies because they center around differing assumptions on the nature of the computational environment. The author describes some of the inconsistencies that arose during the development of Chrysalis++ and suggests possible solutions.< >

关键词： Computer languages Computer aided software engineering Operating systems Parallel programming Microprocessors Computer science object oriented programming object oriented modeling Parallel processing Runtime

来源：评论

学校读者我要写书评

暂无评论

Parallel object-oriented programming in SYMPAL

Parallel object-oriented programming in SYMPAL

引用

Israel conference on Computer systems and Software Engineering

作者： I. Danieli S. Cohen Computer Sciences Department School of Mathematical Sciences Raymond & Beverly Sackler Faculty of Exact Sciences Tel-Aviv University Israel

An object-oriented programming model in a parallel system is presented. It is designed for modelling, describing and solving a wide variety of AI problems. AI applications, in particular, deal with knowledge-bases and a lot of problems have parallel solutions. Therefore, while parallel computers are becoming widespread, there is a need for a rich language that enables the exploitation of parallelism and the achievement of faster solutions. SYMPAL, a parallel symbolic environment and language, that evolved from Common-Lisp, is described, with an emphasis on the POOP (parallel object-oriented programming) style.< >

关键词： object oriented programming object oriented modeling Concurrent computing Artificial intelligence Parallel processing Printers Application software Timing Delay Load management

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：