检索结果-内蒙古大学图书馆

12th International Conference on Practical Applications of Computational Biology and Bioinformatics (PACBB)

作者： Salguero, Alberto G. Capel, Manuel I. Tomeu, Antonio J. Univ Cadiz Cadiz Spain Univ Granada Granada Spain

ISBN: (纸本)9783319987026;9783319987019

"In silico" experimentation allows us to simulate the effect of different therapies by handling model parameters. Although the computational simulation of tumors is currently a well-known technique, it is however possible to contribute to its improvement by parallelizing simulations on computer systems of many and multi-cores. This work presents a proposal to parallelize a tumor growth simulation that is based on cellular automata by partitioning of the data domain and by dynamic load balancing. The initial results of this new approach show that it is possible to successfully accelerate the calculations of a known algorithm for tumor-growth.

关键词： Cellular automaton High performance computing Mathematical oncology Tumoral growth simulation parallel programming Speedup

来源：评论

学校读者我要写书评

暂无评论

Dataflow Execution of Hierarchically Tiled Arrays 25th

Dataflow Execution of Hierarchically Tiled Arrays

引用

25th International Conference on parallel and Distributed Computing (Euro-Par)

作者： Yang, Chih-Chieh Pichel, Juan C. Padua, David A. IBM Corp Thomas J Watson Res Ctr POB 704 Yorktown Hts NY 10598 USA Univ Santiago de Compostela CiTIUS Santiago De Compostela Spain Univ Illinois Urbana IL 61801 USA

ISBN: (纸本)9783030294007;9783030293994

As the parallelism in high-performance supercomputers continues to grow, new programming models become necessary to maintain programmer productivity at today's levels. Dataflow is a promising execution model because it can represent parallelism at different granularity levels and to dynamically adapt for efficient execution. The downside is the low-level programming interface inherent to dataflow. We present a strategy to translate programs written in Hierarchically Tiled Arrays (HTA) to the dataflow API of Open Community Runtime (OCR) system. The goal is to enable program development in a convenient notation and at the same time take advantage of the benefits of a dataflow runtime system. Using HTA produces more comprehensive codes than those written using the dataflow runtime programming interface. Moreover, the experiments show that, for applications with high asynchrony and sparse data dependences, our implementation delivers superior performance than OpenMP using parallel for loops.

关键词： parallel programming Dataflow High-level programming abstraction parallel algorithm

来源：评论

学校读者我要写书评

暂无评论

Application of parallel Genetic Algorithm for Model-Based Gaussian Cluster Analysis 1

引用

9th International Conference on Innovations in Bio- Inspired Computing and Applications (IBICA) / 7th World Congress on Information and Communication Technologies (WICT)

作者： Laurinec, Peter Jarabek, Tomas Lucka, Maria Slovak Univ Technol Bratislava Fac Informat & Informat Technol Ilkovicova 2 Bratislava 84216 Slovakia

ISBN: (数字)9783030166816

ISBN: (纸本)9783030166816;9783030166809

Proposed paper presents a new model-based Gaussian clustering method and defines new optimization criteria for model-based clustering, which are used as fitness functions in genetic algorithm. These optimization criteria are based on different properties of covariance matrices. The proposed model-based Gaussian clustering method is compared with the well-known K-Means method that is solved by genetic algorithm or by Particle Swarm Optimization method. Our method achieves higher similarity between real classification and computed clustering results on all six presented real-world datasets. Because of the high computational requirements of the used methods we have focused on their parallelization. Due to the chosen parallel computer architecture we have combined both MPI and OpenMP programing interfaces. We show that parallelization of the proposed method is very effective and scalable on many execution units.

关键词： Model-based Gaussian cluster analysis Genetic algorithm parallel programming K-Means

来源：评论

学校读者我要写书评

暂无评论

Proving Data Race Freedom in Task parallel Programs Using a Weaker Partial Order 19

Proving Data Race Freedom in Task Parallel Programs Using a ...

引用

19th International Conference on Formal Methods in Computer Aided Design (FMCAD)

作者： Ogles, Benjamin Aldous, Peter Mercer, Eric Brigham Young Univ Provo UT 84601 USA

ISBN: (纸本)9780983567899

Task parallel programming models such as Habanero Java help developers write idiomatic parallel programs and avoid common errors. Data race freedom is a desirable property for task parallel programs but is difficult to prove because every possible execution of the program must be considered. A partial order over events of an observed program execution induces an equivalence class of executions that the program may also produce. The Does-not-Commute (DC) relation is an efficiently computable partial order used for data race detection. As a relatively weak partial order, the DC relation can represent relatively large equivalence classes of program executions. However, some of these executions may be infeasible, thus leading to false data race reports. The contribution of this paper is a mechanized proof that the DC relation is actually sound for commonly used task parallel programming models. Sound means that the first data race identified by the DC relation is guaranteed to be a real data race. A prototype analysis in the Java Pathfinder model checker shows that the DC relation can significantly reduce the number of explored states required to prove data race freedom in Habanero Java programs. In this application, the search for data race using the DC relation is both sound and complete.

关键词： History Task analysis Java Instruction sets Data models Semantics parallel programming

来源：评论

学校读者我要写书评

暂无评论

Theoretical and Practical Approaches for Teaching parallel Code Correctness 26

Theoretical and Practical Approaches for Teaching Parallel C...

引用

26th International Conference on High Performance Computing, Data and Analytics (HiPCW)

作者： Redondo, Carlos Arora, Ritu Trung Nguyen Ba Univ Texas Austin Austin TX 78712 USA Univ Texas Austin Texas Adv Comp Ctr Austin TX 78712 USA Univ Massachusetts Amherst MA 01003 USA

ISBN: (纸本)9781728148946

The introductory-level courses on parallel programming, typically, do not cover the topic of code correctness. Often, students learn about the logical errors in parallel programs and troubleshoot them through trial and error, and spend a significant amount of time and effort in the process. A systematic pedagogical approach to teaching parallel code correctness is therefore needed to enhance the productivity of students and instructors. In this paper, we describe some theoretical and practical approaches that can be adopted for assessing and teaching parallel code correctness. The theoretical approaches include using formal methods (e.g., Petri nets and Hoare Logic). We apply these approaches on the test cases discussed in this paper. The practical approach involves teaching code correctness through demonstrations. For enabling this, we have not only curated a repository of parallel programs with commonly made logical errors but have also added a high-level interface on top of the repository for quickly comparing fixed and incorrect versions of the sample code in the repository, seeing the explanation text on the errors, and searching the repository on the basis of the causes and symptoms of logical errors. The work presented in this paper can potentially motivate the instructors in including the content on code correctness in their parallel programming courses and trainings.

关键词： formal verification Hoare logic Petri nets mathematical modeling parallel programming CUDA MPI OpenMP

来源：评论

学校读者我要写书评

暂无评论

Hardware-Accelerated Energy-Efficient Synchronization and Communication for Ultra-Low-Power Tightly Coupled Clusters 22

Hardware-Accelerated Energy-Efficient Synchronization and Co...

引用

22nd Design, Automation and Test in Europe Conference and Exhibition (DATE)

作者： Glaser, Florian Haugou, Germain Rossi, Davide Huang, Qiuting Benini, Luca Swiss Fed Inst Technol Integrated Syst Lab Zurich Switzerland Univ Bologna Elect Elect & Informat Engn Bologna Italy

ISBN: (纸本)9783981926323

parallel ultra low power computing is emerging as an enabler to meet the growing performance and energy efficiency demands in deeply embedded systems such as the end-nodes of the internet-of-things (IoT). The parallel nature of these systems however adds a significant degree of complexity as processing elements (PEs) need to communicate in various ways to organize and synchronize execution. Naive implementations of these central and non-trivial mechanisms can quickly jeopardize overall system performance and limit the achievable speedup and energy efficiency. To avoid this bottleneck, we present an event-based solution centered around a technology-independent, light-weight and scalable (up to 16 cores) synchronization and communication unit (SCU) and its integration into a shared-memory multicore cluster. Careful design and tight coupling of the SCU to the data interfaces of the cores allows to execute common synchronization procedures with a single instruction. Furthermore, we present hardware support for the common barrier and lock synchronization primitives with a barrier latency of only eleven cycles, independent of the number of involved cores. We demonstrate the efficiency of the solution based on experiments with a post-layout implementation of the multicore cluster in a 22 nm CMOS process where the SCU constitutes less than 2% of area overhead. Our solution supports parallel sections as small as 100 or 72 cycles with a synchronization overhead of just 10 %, an improvement of up to 14x or 30x with respect to cycle count or energy, respectively, compared to a test-and-set based implementation.

关键词： Synchronization Multicore processing Hardware Clocks Software parallel programming

来源：评论

学校读者我要写书评

暂无评论

A tale of lock-free agents: towards Software Transactional Memory in parallel Agent-Based Simulation

引用

COMPLEX ADAPTIVE SYSTEMS MODELING 2019年第1期7卷 1-28页

作者： Thaler, Jonathan Siebers, Peer-Olaf Univ Nottingham 7301 Wollaton Rd Nottingham NG8 1BB England

With the decline of Moore's law and the ever increasing availability of cheap massively parallel hardware, it becomes more and more important to embrace parallel programming methods to implement Agent-Based Simulations (ABS). This has been acknowledged in the field a while ago and numerous research on distributed parallel ABS exists, focusing primarily on parallel Discrete Event Simulation as the underlying mechanism. However, these concepts and tools are inherently difficult to master and apply and often an excess in case implementers simply want to parallelise their own, custom agent-based model implementation. However, with the established programming languages in the field, Python, Java and C++, it is not easy to address the complexities of parallel programming due to unrestricted side effects and the intricacies of low-level locking semantics. Therefore, in this paper we propose the use of a lock-free approach to parallel ABS using Software Transactional Memory (STM) in conjunction with the pure functional programming language Haskell, which in combination, removes some of the problems and complexities of parallel implementations in imperative approaches. We present two case studies, in which we compare the performance of lock-based and lock-free STM implementations in two different well known Agent-Based Models, where we investigate both the scaling performance under increasing number of CPU cores and the scaling performance under increasing number of agents. We show that the lock-free STM implementations consistently outperform the lock-based ones and scale much better to increasing number of CPU cores both on local hardware and on Amazon EC. Further, by utilizing the pure functional language Haskell we gain the benefits of immutable data and lack of unrestricted side effects guaranteed at compile-time, making validation easier and leading to increased confidence in the correctness of an implementation, something of fundamental importance and benefit in paral

关键词： Agent-Based Simulation Software Transactional Memory parallel programming Haskell

来源：评论

学校读者我要写书评

暂无评论

HPC Education for domain scientists: An Indian experience and perspective 26

HPC Education for domain scientists: An Indian experience an...

引用

26th International Conference on High Performance Computing, Data and Analytics (HiPCW)

作者： Shenoi, V. Venkatesh Shah, Vaishali Joshi, Sandeep K. HPC Infrastruct & Ecosyst Ctr Dev Adv Comp Pune Maharashtra India Savitribai Phule Pune Univ Interdisciplinary Sch Sci Comp Pune Maharashtra India

ISBN: (纸本)9781728148946

We review the HPC training avenues for masters, Ph.D. students, postdocs and young faculty in Indian universities and research institutes. Their interest in HPC arises from their need to use it for research in their scientific domain having done some background courses in programming with Fortran/C/C++ and numerical methods. Very few Indian educational institutes offer a course/courses in parallel programming and the course is mostly offered in some of the high ranking engineering institutes. Even in the stream of Computer Science, the course if at all offered is an elective course. The non-Computer Science domain users that makeup as the majority of users of HPC in India do not possess sufficient background in Computer Science and gear up as HPC users mostly through self-study or short term training workshops that address their requirements. However, such training programmes are not a regular activity in India and cater to a broad audience with a varied level of aptitude in computing, programming, and simulation via mathematical modelling. In this paper, we discuss the current HPC education scenario and propose an education strategy to broaden the base of HPC users and to help researchers at all levels to effectively use HPC for their academic research and development work.

关键词： High Performance Computing HPC parallel programming scientific computing training course

来源：评论

学校读者我要写书评

暂无评论

Case Study: Using Project Based Learning to Develop parallel Programing and Soft Skills 33

Case Study: Using Project Based Learning to Develop Parallel...

引用

33rd IEEE International parallel and Distributed Processing Symposium (IPDPS)

作者： Younis, Awad A. Sunderraman, Rajshekhar Metzler, Mike Bourgeois, Anu G. Georgia State Univ Atlanta GA 30303 USA

ISBN: (纸本)9781538655559

In today's environment, where every computer, including cell phones, are multicore, it is essential that students develop parallel programming skills. It remains a challenge to develop effective techniques for teaching parallel programing skills. Another challenge is finding time within already packed lectures to cover additional material. To that end, we investigate the effectiveness of using Project Based Learning (PBL) to teach parallel programming skills early in the curriculum by developing and incorporating a PBL module into CSc 3210 (Computer Organization and programming). This is a core course taken by all computer science majors and is a prerequisite to many of our senior-level classes. In our case study, 124 students are organized into 26 diverse groups, with four or five students per group, and assigned five project assignments, each of two-weeks duration. Given a Raspberry PI, students will explore its multicore architecture and create programs for shared memory parallelism using OpenMP and C language. Our results show that incorporating this PBL module has a significant and direct effect on the student's growth in parallel programming skills. As a side benefit, we also show that there is a direct improvement on a student's personal growth in terms soft skills, which is essential in the professional development and success in the workplace. By having students experience PBL in an early class, close to the midpoint of the academic program, it can serve as a mini-capstone project. Furthermore, students can collaboratively learn by themselves (through teamwork) and apply the fundamentals of parallel programming skills without the need for separate lectures, labs, or workshops.

关键词： parallel Computing parallel programming OpenMP Shared Memory Project Based Learning Raspberry PI Soft Skills Teamwork Technology

来源：评论

学校读者我要写书评

暂无评论

Novice-Friendly Multi-Armed Robotics programming 2

Novice-Friendly Multi-Armed Robotics Programming

引用

2nd IEEE/ACM International Workshop on Robotics Software Engineering (RoSE)

作者： Ritschel, Nico Holmes, Reid Garcia, Ronald Shepherd, David C. Univ British Columbia Dept Comp Sci Vancouver BC Canada ABB Corp Res Raleigh NC USA

ISBN: (纸本)9781728122496

Collaborative robots are being applied in a growing number of usage scenarios, but their adoption is slowed down by the high complexity of robot programming. As previous prototype studies have shown, block-based programming environments can enable novice or end users to program industrial single-armed robots. Some existing block-based tools support parallel programming and therefore show potential to be used for multi-armed robot programming as well. We analyze their designs and argue how improved abstractions and visualizations could make multi-armed parallelism accessible to novice users. Based on this analysis, we then extract a list of features that a block-based environment designed for multi-armed robot programming should provide. Finally, we present our design vision for a novel programming environment for two-armed robots, show how it provides these features and discuss how it can enable both novices and experienced intermediate users to perform parallelized programming tasks.

关键词： Application programming interfaces programming environments Robot programming parallel 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：