This paper considers the current state of softwareengineering for parallelsystems. A review of existing approaches and techniques identifies inadequacies. Recent work on design, verification and automated support is...
详细信息
This workshop provides a forum for an overview, project presentations, and discussion of the research fostered and funded initially by the NSF Next Generation software (NGS) Program, and more recently through the Adva...
详细信息
An important concept for indivisible updates in parallel computing are atomic operations. For most architectures, they also provide ordering guarantees, which in practice can hurt performance. For associative and comm...
详细信息
ISBN:
(纸本)9781728112466
An important concept for indivisible updates in parallel computing are atomic operations. For most architectures, they also provide ordering guarantees, which in practice can hurt performance. For associative and commutative updates, in this paper we present software buffering techniques that overcome the problem of ordering by combining multiple updates in a temporary buffer and by prefetching addresses before updating them. As a result, our buffering techniques reduce contention and avoid unnecessary ordering constraints, in order to increase the amount of memory parallelism. We evaluate our techniques in different scenarios, including applications like histogram and graph computations, and reason about the applicability for standard systems and multi-socket systems.
In order to build a flexible and configurable software architecture to realize enterprise business requirements agilely for distributing environment, a methodology called BSAD (Business-driven Service-oriented Applica...
详细信息
Computing the Hierarchical Equations of Motion (HEOM) is by itself a challenging problem, and so is writing portable production code that runs efficiently on a variety of architectures while scaling from PCs to superc...
详细信息
ISBN:
(纸本)9781538655559
Computing the Hierarchical Equations of Motion (HEOM) is by itself a challenging problem, and so is writing portable production code that runs efficiently on a variety of architectures while scaling from PCs to supercomputers. We combined both challenges to push the boundaries of simulating quantum systems, and to evaluate and improve methodologies for scientific softwareengineering. Our contributions are threefold: We present the first distributed memory implementation of the HEOM method (DM-HEOM), we describe an interdisciplinary development workflow, and we provide guidelines and experiences for designing distributed, performance-portable HPC applications with MPI3, OpenCL and other state-of-the-art programming models. We evaluated the resulting code on multi- and many-core CPUs as well as GPUs, and demonstrate scalability on a Cray XC40 supercomputer for the PS I molecular light harvesting complex.
Traditional softwaredistributed shared memory (SDSM) systems modify the semantics of a real hardware shared memory system by relaxing the coherence semantic and by limiting the memory regions that are actually shared...
详细信息
Traditional softwaredistributed shared memory (SDSM) systems modify the semantics of a real hardware shared memory system by relaxing the coherence semantic and by limiting the memory regions that are actually shared. These semantic modifications are done to improve performance of the applications using it. In this paper. we will show that a SDSM system that behaves like a real shared memory system (without the afore-mentioned relaxations) can also be used to execute OpenMP applications and achieve similar speedups as the ones obtained by traditional SDSM systems. This performance can be achieved by encouraging the cooperation between the SDSM and the OpenMP runtime instead of relaxing the semantics of the shared memory. In addition, techniques like boundaries alignment and page presend are demonstrated as very useful to overcome the limitations of the current SDSM systems. (c) 2005 Elsevier Inc. All rights reserved.
This paper first reviews the application of computing technology in electric power system with emphasis on the evolution of the software architecture. And then proposes a software architecture to implement power syste...
详细信息
ISBN:
(纸本)0769525547
This paper first reviews the application of computing technology in electric power system with emphasis on the evolution of the software architecture. And then proposes a software architecture to implement power system operation and computing. As power systems have become more complicated, software tools and simulations have played a more important role. The operation and control of the power system is actually becoming a huge data-intensive, information-intensive, communication-intensive and computing-intensive application. Power systems rely more heavily on computerized communications and control. Therefore,. system security has. become increasingly dependent on protecting the integrity of the associated information systems. Grid computing is generally regarded as a software technology to fully use the spare computing resources. However, the philosophy of grid computing can be used in engineering case to play an important role in power system distributed monitoring, control and distributedparallel computing. This paper proposes a software architecture, which depends on grid computing for hardware support and agent technology for software support, to seamlessly integrate the dispersed computing resources to implement high-performance operation and computing in electric power system. Some comparative test cases are studied, and the test results show that the combination of agent and grid computing can enhance the performance of the distributed computing system.
Over the academic year 2022-23, we discussed the teaching of software performance engineering with more than a dozen faculty across North America and beyond. Our outreach was centered on research-focused faculty with ...
详细信息
ISBN:
(纸本)9798350364613;9798350364606
Over the academic year 2022-23, we discussed the teaching of software performance engineering with more than a dozen faculty across North America and beyond. Our outreach was centered on research-focused faculty with an existing interest in this course material. These discussions revealed an enthusiasm for making software pertimmance engineering a more prominent part of a curriculum for computer scientists and engineers. Here, we discuss how MIT's longstanding efforts in this area may serve as a launching point for community development of a software performance engineering curriculum, challenges in and solutions for providing the necessary infrastructure to universities, and future directions.
This paper presents the specification and implementation procedure using a microcomputer network based autonomous distributed control architecture for industrial multirobot systems. The procedure is based on the conce...
详细信息
This paper presents the specification and implementation procedure using a microcomputer network based autonomous distributed control architecture for industrial multirobot systems. The procedure is based on the concept of data flow network controlled by communicating sequential processes to perform coordinated tasks. Robots and other computerized industrial devices such as conveyors and manufacturing machines are defined as object-oriented Petri nets. A modular and hierarchical approach is adopted to define a set of Petri net type diagrams which represent concurrent activities of control processes for such devices. Asynchronous and synchronous interactions are modelled by places and transitions, respectively, in global process interaction nets. The control software is implemented on a computer network using Inmos transputers with true parallel processing and message passing primitives efficiently handled in hardware. Petri net based models are directly and efficiently transformed to corresponding codes in occam, the high level parallel. programming language defined for the transputer.
We employ probabilistic causality analysis to study the performance data of 301 students from the upper-level undergraduate parallel programming class at the University of Central Florida. To our surprise, we discover...
详细信息
ISBN:
(纸本)9781538655559
We employ probabilistic causality analysis to study the performance data of 301 students from the upper-level undergraduate parallel programming class at the University of Central Florida. To our surprise, we discover that good performance in our lower-level undergraduate programming CS-1 and CS-II classes is not a significant causal factor that contributed to good performance in our parallel programming class. On the other hand, good performance in systems classes like Operating systems, Information Security, Computer Architecture, Object Oriented software and systemssoftware coupled with good performance in theoretical classes like Introduction to Discrete Structures, Artificial Intelligence and Discrete Structures-II are strong indicators of good performance in our upper-level undergraduate parallel programming class. We believe that such causal analysis may be useful in identifying whether parallel and distributed computing concepts have effectively penetrated the lower-level computer science classes at an institution.
暂无评论