检索结果-内蒙古大学图书馆

Improved OpenCL-Based Implementation of Social Field Pedestrian Model

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS 2020年第7期21卷 2828-2839页

作者： Yu, Bin Zhu, Ke Wu, Kaiteng Zhang, Michael Tongji Univ Key Lab Rd & Traff Engn Minist Educ Shanghai 201804 Peoples R China Tongji Univ Coll Transportat Engn Shanghai 201804 Peoples R China Univ Calif Davis Dept Civil & Environm Engn Davis CA 95616 USA

Two aspects of improvements are proposed for the OpenCL-based implementation of the social field pedestrian model. In the aspect of algorithm, a method based on the idea of divide-and-conquer is devised in order to overcome the problem of global memory depletion when fields are of a larger size. This is of importance for the study of finer pedestrian walking behavior, which usually requires larger fields. In the aspect of computation, the OpenCL heterogeneous framework is thoroughly studied. Factors that may affect the numerical efficiency are evaluated, with regarding to the social field model previously proposed. This includes usage of local memory, deliberate patch of data structures for avoidance of bank conflicts, and so on. Experiments disclose that the numerical efficiency is brought to an even higher level. Compared with the CPU model and the previous GPU model, the present GPU model can be at most 71.56 and 13.3 times faster, respectively.

关键词： Computational modeling Graphics processing units Force Numerical models Legged locomotion Computer architecture Intelligent transportation systems Algorithm crowd dynamics heterogeneous parallel computing OpenCL pedestrian flow

来源：评论

学校读者我要写书评

暂无评论

Assessing Application Efficiency and Performance Portability in Single-Source Programming for heterogeneous parallel Systems

引用

INTERNATIONAL JOURNAL OF parallel PROGRAMMING 2023年第1期51卷 61-82页

作者： Ernstsson, August Griebler, Dalvan Kessler, Christoph Linkoping Univ Dept Comp & Informat Sci PELAB Linkoping Sweden Pontif Catholic Univ Rio Grande do Sul PUCRS Sch Technol Porto Alegre Brazil

We analyze the performance portability of the skeleton-based, single-source multi-backend high-level programming framework SkePU across multiple different CPU-GPU heterogeneous systems. Thereby, we provide a systematic application efficiency characterization of SkePU-generated code in comparison to equivalent hand-written code in more low-level parallel programming models such as OpenMP and CUDA. For this purpose, we contribute ports of the STREAM benchmark suite and of a part of the NAS parallel Benchmark suite to SkePU. We show that for STREAM and the EP benchmark, SkePU regularly scores efficiency values above 80% and in particular for CPU systems, SkePU can outperform hand-written code.

关键词： Algorithmic skeletons parallel efficiency Performance portability heterogeneous parallel computing High-level parallel programming

来源：评论

学校读者我要写书评

暂无评论

T-ReX: Interactive Global Illumination of Massive Models on heterogeneous computing Resources

引用

IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS 2014年第3期20卷 481-494页

作者： Kim, Tae-Joon Sun, Xin Yoon, Sung-Eui Korea Adv Inst Sci & Technol Dept Comp Sci Taejon 305701 South Korea Microsoft Res Asia Beijing 100080 Peoples R China

We propose several interactive global illumination techniques for a diverse set of massive models. We integrate these techniques within a progressive rendering framework that aims to achieve both a high rendering throughput and an interactive responsiveness. To achieve a high rendering throughput, we utilize heterogeneous computing resources consisting of CPU and GPU. To reduce expensive data transmission costs between CPU and GPU, we propose to use separate, decoupled data representations dedicated for each CPU and GPU. Our representations consist of geometric and volumetric parts, provide different levels of resolutions, and support progressive global illumination for massive models. We also propose a novel, augmented volumetric representation that provides additional geometric resolutions within our volumetric representation. In addition, we employ tile-based rendering and propose a tile ordering technique considering visual perception. We have tested our approach with a diverse set of large-scale models including CAD, scanned, simulation models that consist of more than 300 million triangles. By using our methods, we are able to achieve ray processing performances of 3 M similar to 20 M rays per second, while limiting response time to users within 15 similar to 67 ms. We also allow dynamic modifications of light, and interactive setting of materials, while efficiently supporting novel view rendering.

关键词： Massive models ray tracing photon mapping global illumination heterogeneous parallel computing voxels compression

来源：评论

学校读者我要写书评

暂无评论

Improving the scalability of hyperspectral imaging applications on heterogeneous platforms using adaptive run-time data compression

引用

COMPUTERS & GEOSCIENCES 2010年第10期36卷 1283-1291页

作者： Plaza, Antonio Plaza, Javier Paz, Abel Univ Extremadura Dept Technol Comp & Commun Hyperspectral Comp Lab E-10071 Caceres Spain

Latest generation remote sensing instruments (called hyperspectral imagers) are now able to generate hundreds of images, corresponding to different wavelength channels, for the same area on the surface of the Earth. In previous work, we have reported that the scalability of parallel processing algorithms dealing with these high-dimensional data volumes is affected by the amount of data to be exchanged through the communication network of the system. However, large messages are common in hyperspectral imaging applications since processing algorithms are pixel-based, and each pixel vector to be exchanged through the communication network is made up of hundreds of spectral values. Thus, decreasing the amount of data to be exchanged could improve the scalability and parallel performance. In this paper, we propose a new framework based on intelligent utilization of wavelet-based data compression techniques for improving the scalability of a standard hyperspectral image processing chain on heterogeneous networks of workstations. This type of parallel platform is quickly becoming a standard in hyperspectral image processing due to the distributed nature of collected hyperspectral data as well as its flexibility and low cost. Our experimental results indicate that adaptive lossy compression can lead to improvements in the scalability of the hyperspectral processing chain without sacrificing analysis accuracy, even at sub-pixel precision levels. (C) 2010 Elsevier Ltd. All rights reserved.

关键词： heterogeneous parallel computing Adaptive run-time data compression Wavelet transform Hyperspectral imaging Remote sensing

来源：评论

学校读者我要写书评

暂无评论

A GPU-Accelerated automated multilevel substructuring method for modal analysis of structures

引用

COMPUTERS & STRUCTURES 2024年 305卷

作者： Wang, Guidong Wang, Yujie Chen, Zeyu Wang, Feiqi Li, She Cui, Xiangyang Hunan Univ State Key Lab Adv Design & Mfg Technol Vehicle Changsha 410082 Peoples R China Hunan Maixi Software Co Ltd Changsha 410082 Peoples R China

In this work, a novel GPU-accelerated heterogeneous method for the automated multilevel substructuring method(HAMLS) is presented for dealing large finite element models in structural dynamics. Different parallel modes based on node, subtree, and eigenpair have been developed in the solution steps of AMLS to achieve a heterogeneous strategy. First, a new data management method is designed during the model transformation phase to eliminate the determinacy race in the parallel strategy of the separator tree. Considering the distribution characteristics of the nodes in the separator tree and the dependence of node tasks, a load balancing heterogeneous parallel strategy is designed to take full advantage of hosts and devices. By developing an adaptive batch processing program for solving eigenvectors during the back transformation phase, the overheads of launching kernels, as well as the GPU memory requirements, can be reduced by several orders of magnitude. Several numerical examples have been employed to validate the efficiency and practicality of the novel GPU-accelerated heterogeneous strategy. The results demonstrate that the computational efficiency of the novel strategy using one GPU can increase to 3.0x that of the original parallel AMLS method when 16 CPU threads are used.

关键词： AMLS method heterogeneous parallel computing Finite element analysis Eigenvalue problem

来源：评论

学校读者我要写书评

暂无评论

Efficient Utilization of Multi-Threading parallelism on heterogeneous Systems for Sparse Tensor Contraction

引用

IEEE TRANSACTIONS ON parallel AND DISTRIBUTED SYSTEMS 2024年第6期35卷 889-900页

作者： Xiao, Guoqing Yin, Chuanghui Chen, Yuedan Duan, Mingxing Li, Kenli Hunan Univ Coll Comp Sci & Elect Engn Changsha 410082 Peoples R China Hunan Univ Shenzhen Inst Shenzhen 518063 Peoples R China Cent South Univ Big Data Inst Changsha 410083 Peoples R China

Many fields of scientific simulation, such as chemistry and condensed matter physics, are increasingly eschewing dense tensor contraction in favor of sparse tensor contraction. In this work, we center around binary sparse tensor contraction (SpTC) which has the challenges of index matching and accumulation. To address these difficulties, we present GSpTC, an efficient element-wise SpTC framework on CPU-GPU heterogeneous systems. GSpTC first introduces a fine-grained partitioning strategy based on element-wise tensor contraction. By analyzing and selecting appropriate dimension partitioning strategies, we can efficiently utilize the multi-threading parallelism on GPUs and optimize the overall performance of GSpTC. In particular, GSpTC leverages multi-threading parallelism on GPUs for the contraction phase and merging phase, which greatly accelerates the computation phase in sparse tensor contraction computations. Furthermore, GSpTC employs parallel pipeline technology to hide the data transmission time between the host and the device, further enhancing its performance. As a result, GSpTC achieves an average performance improvement of 267% compared to the previous state-of-the-art framework Sparta.

关键词： Tensors parallel processing Performance evaluation Graphics processing units Sparse matrices Partitioning algorithms Sorting heterogeneous parallel computing multi-threading parallelism optimization parallel pipeline sparse tensor contraction

来源：评论

学校读者我要写书评

暂无评论

parallel PROCESSING OF REMOTELY SENSED HYPERSPECTRAL IMAGES ON heterogeneous NETWORKS OF WORKSTATIONS USING HETEROMPI

引用

INTERNATIONAL JOURNAL OF HIGH PERFORMANCE computing APPLICATIONS 2008年第4期22卷 386-407页

作者： Valencia, David Lastovetsky, Alexey O'Flynn, Maureen Plaza, Antonio Plaza, Javier Univ Extremadura Tech Sch Caceres Dept Technol Comp & Commun E-10071 Caceres Spain Univ Coll Dublin Sch Comp Sci & Informat Heterogeneous Comp Lab Dublin 4 Ireland

The development of efficient techniques for transforming massive volumes of remotely sensed hyperspectral data into scientific understanding is critical for space-based Earth science and planetary exploration. Although most available parallel processing strategies for information extraction and mining from hyperspectral imagery assume homogeneity in the underlying computing platform, heterogeneous networks of computers (HNOCs) have become a promising cost-effective solution, expected to play a major role in many on-going and planned remote sensing missions. In this paper, we develop a new morphological parallel algorithm for hyperspectral image classification using HeteroMPI, an extension of MPI for programming high-performance computations on HNOCs. The main idea of HeteroMPI is to automate and optimize the selection of a group of processes that executes a heterogeneous algorithm faster than any other possible group in a heterogeneous environment. In order to analyze the impact of many-to-one ( gather) communication operations introduced by our proposed algorithm, we resort to a recently proposed collective communication model. The parallel algorithm is validated using two heterogeneous clusters at University College Dublin and a massively parallel Beowulf cluster at NASA's Goddard Space Flight Center.

关键词： heterogeneous parallel computing hyperspectral image processing HeteroMPI mathematical morphology performance evaluation

来源：评论

学校读者我要写书评

暂无评论

An OpenCL Implementation of Ellipsoidal Harmonics 8th

An OpenCL Implementation of Ellipsoidal Harmonics

引用

8th Hotine-Marussi Symposium on Mathematical Geodesy

作者： Nesvadba, Otakar Holota, Petr Land Survey Off Sidlistem 9 Prague 18200 Czech Republic Res Inst Geodesy Topog & Cartog Prague East Zdiby 25066 98 Czech Republic

ISBN: (纸本)9783319305301;9783319245485

The technology progress today makes it possible to treat most of the problems of physical geodesy by means of numerical arrangements hardly imaginable earlier. Nevertheless, considering an evaluation of spheroidal (spherical and ellipsoidal) harmonic functions in our typical tasks, we still observe a huge performance gap between our demands and capabilities of common CPUs. Methods used for calculating associated Legendre functions are mostly recursive and thus sequential. Therefore, it is challenging, but feasible, to arrange the processing of Legendre functions in a way that reduces memory utilisation and admits massive parallelism. Following this aim, we developed a streaming-parallel algorithm for computing oblate spheroidal harmonic functions and their derivatives. The algorithm is free of assumptions concerning the function arguments, maximal degree/order or number of computation points and can be utilised on any data type, like a vector or scalar float, double or even integer numbers. Besides, it solves floating-point issues in the numerical treatment of Legendre functions. We demonstrate its Open computing Language (OpenCL) implementation on a general-purpose graphics processing unit (GPGPU), which is ideal for its inexpensive computational power of some TFlops. Added performance benchmarks lead to the conclusion that our implementation on a single GPGPU device substantially outperforms recent multi-core CPUs, free of any precision penalty. Furthermore, thanks to the OpenCL standard, we can benefit from an excellent portability and scalability over heterogeneous parallel platforms. Let us note finally, that the topic presented is a matter of importance in many other application fields, not only in physical geodesy.

关键词： Associated Legendre functions Clenshaw summation heterogeneous parallel computing Oblate spheroidal harmonics OpenCL

来源：评论

学校读者我要写书评

暂无评论

Closed-form Solutions for Dense Matrix-Matrix Multiplication on heterogeneous Platforms using Divisible Load Analysis 26

Closed-form Solutions for Dense Matrix-Matrix Multiplication...

引用

26th Euromicro International Conference on parallel, Distributed, and Network-Based Processing (PDP)

作者： Barlas, Gerssimos El Hiny, Lamees Amer Univ Sharjah Comp Sci & Engn Dept Coll Engn Sharjah U Arab Emirates

ISBN: (纸本)9781538649756

In this paper we analytically solve the partitioning problem for performing matrix multiplication on a cluster of heterogeneous multicore machines, equipped with an accelerator, typically a GPU. We derive closed-form solutions that not only solve the problem in an exact manner, but they also allow for predictive analysis that can guide system design. Our work allows an optimum partitioning to be calculated in linear time with respect to the number of cores in the system. The static partitioning afforded by our Divisible Load Theory (DLT) based analysis, minimizes communication overhead and improves efficiency. Our work leverages existing optimized Dense Linear Algebra (DLA) libraries, such as cuBLAS and BLAS, which translates to an easy deployment that can readily exploit state-of-the-art tools. A comparison study concludes the paper, highlighting the beneficial effect of our partitioning approach.

关键词： GPU computing dense linear algebra heterogeneous parallel computing hybrid numerical computation divisible load theory

来源：评论

学校读者我要写书评

暂无评论

Performance Optimization of Vision Apps on Mobile Application Processor

Performance Optimization of Vision Apps on Mobile Applicatio...

引用

20th International Conference on Systems, Signals and Image Processing (IWSSIP)

作者： Cheng, Kwang-Ting (Tim) Yang, Xin Wang, Yi-Chu Univ Calif Santa Barbara Dept Elect & Comp Engn Santa Barbara CA 93106 USA

ISBN: (纸本)9781479909414;9781479909421

Optimizing performance of compute-intensive vision apps running on mobile application processor (AP) is critical to satisfactory experience for smartphone and tablet users. Most existing vision algorithms were primarily designed and implemented for desktop and server platforms. Porting them to a mobile platform without adapting the algorithms to account for the platform's limitations would cause serious algorithm-hardware mismatches, yielding unnecessary runtime degradation. Modern mobile AP, which integrates multi-core CPUs, GPUs and other special-purpose accelerators, offers multiple options of porting vision apps to various computing cores. To develop an optimized implementation for a vision app, it is necessary to understand the potential mismatches for better algorithm adaptation and optimized mapping of the algorithm to a handheld platform. In this paper, we identify mismatches and propose adaptation guidelines for three different porting strategies: porting an algorithm to 1) a mobile CPU, 2) a mobile GPU, and 3) mobile heterogeneous multi-cores ( mobile CPUs+GPUs). For each strategy, we illustrate the adaptation/porting guidelines using an exemplar vision task. Experimental results demonstrate that with proper adaptation following the proposed guidelines, we could achieve a significant speedup with little accuracy drop.

关键词： mobile application processor algorithm adaption GPU acceleration heterogeneous parallel computing

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：