检索结果-内蒙古大学图书馆

IEEE Int Conf on Parallel and Distributed Processing with Applications, Big Data and Cloud Computing, Sustainable Computing and Communications, Social Computing and Networking (ISPA/BDCloud/SocialCom/SustainCom)

作者： Li, Mingfan Jiang, Qingcai Lin, Han An, Hong Univ Sci & Technol China Hefei Peoples R China

ISBN: (纸本)9781728143286

Massively multithreaded gpus achieve considerable performance improvements by thousands of processing cores running simultaneously for compute-intensive applications. However, present parallel programming model suffers significant performance degradation on dependency scenarios, where tasks are assigned into multiple thread blocks and the parallelism is limited by inter-block dependencies. This paper propose Gdarts, a dataflow runtime system for gpu workloads with data dependencies. Employing asynchronous task programming, Gdarts builds Stream Multiprocessors (SM) as standalone workers towards fine-granular and independent kernels. Meanwhile, it designs a two-level task scheduler to adapt gpu's hierarchical memory environment. Such a scheme, leveraging the hybrid memory resources, fulfills requirements for workload balancing across SMs, as well as flexible priority scheduling. Based on the improved controllability and flexible support for kernel scheduling, Gdarts complements the existing gpu with insights of dataflow diagram by two heuristic schedule approaches, node-degree and data-driven priority.

关键词： task-based execution dataflow runtime system schedule gpu programming

来源：评论

学校读者我要写书评

暂无评论

Position control of an acoustic cavitation bubble by reinforcement learning

引用

ULTRASONICS SONOCHEMISTRY 2025年 115卷 107290页

作者： Klapcsik, Kalman Gyires-Toth, Balint Rossello, Juan Manuel Hegedus, Ferenc Budapest Univ Technol & Econ Fac Mech Engn Dept Hydrodynam Syst Egyet Rkp 3 H-1111 Budapest Hungary Budapest Univ Technol & Econ Fac Elect Engn & Informat Dept Telecommun & Media Informat Egyet Rkp 3 H-1111 Budapest Hungary Univ Ljubljana Fac Mech Engn Askerceva 6 Ljubljana 1000 Slovenia

Reinforcement Learning (RL) is employed to develop control techniques for manipulating acoustic cavitation bubbles. This paper presents a proof of concept in which an RL agent is trained to discover a policy that allows precise control of bubble positions within a dual-frequency standing acoustic wave field by adjusting the pressure amplitude values. The agent is rewarded for driving the bubble to a target position in the shortest possible time. The results demonstrate that the agent exploits the nonlinear behaviour of the bubble and, in specific cases, identifies solutions that cannot be addressed using the linear theory of the primary Bjerknes force. The RL agent performs well under domain randomization, indicating that the RL approach generalizes effectively and produces models robust against noise, which could arise in real-world applications.

关键词： Bubble position control Reinforcement learning Bubble dynamics gpu programming

来源：评论

学校读者我要写书评

暂无评论

Quasar - a new programming framework for real-time image/video processing on gpu and CPU 15

Quasar - a new programming framework for real-time image/vid...

引用

Proceedings of the 9th International Conference on Distributed Smart Cameras

作者： Bart Goossens Jonas De Vylder Simon Donné Wilfried Philips Ghent University - TELIN/IPI/iMinds Ghent Belgium

ISBN: (纸本)9781450336819

In this demonstration, we present a new programming framework, Quasar, for heterogeneous programming on CPU and single/multi-gpu. Our programming framework consists of a high-level language that is aimed at relieving the programmer from hardware-related implementation issues that commonly occur in CPU/gpu programming, allowing the programmer to focus on the specification, the design, testing and the improvement of the algorithms. We will demonstrate a real-time multi-camera processing application using our integrated development environment (IDE). The IDE offers various image/video processing-related debugging functions and performance profiling features.

关键词： gpu programming

来源：评论

学校读者我要写书评

暂无评论

ForOpenCL: transformations exploiting array syntax in Fortran for accelerator programming

引用

INTERNATIONAL JOURNAL OF COMPUTATIONAL SCIENCE AND ENGINEERING 2013年第1期8卷 47-57页

作者： Sottile, Matthew J. Rasmussen, Craig E. Weseloh, Wayne N. Robey, Robert W. Quinlan, Daniel Overbey, Jeffrey Galois Inc Portland OR 97204 USA Los Alamos Natl Lab Los Alamos NM 87545 USA Los Alamos Natl Lab Silverton Project Los Alamos NM 87545 USA Los Alamos Natl Lab Eulerian Applicat Grp Los Alamos NM 87545 USA Lawrence Livermore Natl Lab Ctr Adv Sci Comp ROSE Project Livermore CA 94550 USA Natl Ctr Supercomp Applicat Urbana IL 61801 USA

Emerging gpu architectures for high performance computing are well suited to a data-parallel programming model. This paper presents preliminary work examining a programming methodology that provides Fortran programmers with access to these emerging systems. We use array constructs in Fortran to show how this infrequently exploited, standardised language feature is easily transformed to lower-level accelerator code. The transformations in ForOpenCL are based on a simple mapping from Fortran to OpenCL. We demonstrate, using a stencil code solving the shallow-water fluid equations, that the performance of the ForOpenCL compiler-generated transformations is comparable with that of hand-optimised OpenCL code.

关键词： Fortran OpenCL accelerator programming gpu programming code generation

来源：评论

学校读者我要写书评

暂无评论

C-for-Metal: High Performance SIMD programming on Intel gpus 21

C-for-Metal: High Performance SIMD Programming on Intel GPUs

引用

19th IEEE/ACM International Symposium on Code Generation and Optimization (CGO)

作者： Lueh, Guei-Yuan Chen, Kaiyu Chen, Gang Fuentes, Joel Chen, Wei-Yu Fu, Fangwen Jiang, Hong Li, Hongzheng Rhee, Daniel Intel Corp Santa Clara CA 95054 USA

ISBN: (纸本)9781728186139

The SIMT execution model is commonly used for general gpu development. CUDA and OpenCL developers write scalar code that is implicitly parallelized by compiler and hardware. On Intel gpus, however, this abstraction has profound performance implications as the underlying ISA is SIMD and important hardware capabilities cannot be fully utilized. To close this performance gap we introduce C-For-Metal (CM), an explicit SIMD programming framework designed to deliver close-to-the-metal performance on Intel gpus. The CM programming language and its vector/matrix types provide an intuitive interface to exploit the underlying hardware features, allowing fine-grained register management, SIMD size control and cross-lane data sharing. Experimental results show that CM applications from different domains outperform the best-known SIMT-based OpenCL implementations, achieving up to 2.7x speedup on the latest Intel gpu.

关键词： SIMD SIMT gpu programming

来源：评论

学校读者我要写书评

暂无评论

Parallel programming Approaches to Optimize Performance and Energy Consumption on Heterogeneous Computing Systems

Parallel Programming Approaches to Optimize Performance and ...

引用

作者： Bauer, Brian Harvard University

学位级别：A.L.M., Master of Liberal Arts

Parallel programming offers the ability to simultaneously improve the performance and reduce the energy consumption of software running on heterogeneous computing systems. Software developers have long preferred to avoid parallel programming, if possible, for reasons such as perceived difficulty, lack of portability between systems, and the pace of improvement in computer hardware. However, generational changes in computer hardware are now focused on specialized components and increased computational cores, and the continued evolution of these systems places increased emphasis on achieving improvements via the use of these components. This thesis investigates parallel programming techniques that make use of components common to modern heterogeneous systems, and proposes that the difficulty and lack of portability need not be barriers to large improvements. Using a variety of heterogeneous systems, algorithms were implemented and then transformed using multiple cores, SIMD execution units, and gpus. Reductions in execution time ranging from 71-94% and energy consumption of 76-98% were observed, demonstrating the effectiveness of using specialized components for improved performance and reduced energy consumption.

关键词： Energy consumption gpu programming Heterogeneous systems Parallel programming SIMD programming Software engineering

来源：评论

学校读者我要写书评

暂无评论

Providing Source Code Level Portability Between CPU and gpu with MapCG

引用

Journal of Computer Science & Technology 2012年第1期27卷 42-56页

作者： Chun-Tao Hong De-Hao Chen Yu-Bei Chen Wen-Guang Chen Wei-Min Zheng Hai-Bo Lin Department of Computer Science and Technology Tsinghua UniversityBeijing 100084China Department of Electronic Engineering Tsinghua UniversityBeijing 100084China IBM China Research Lab Beijing 100094China

Graphics processing units （gpu） have taken an important role in the general purpose computing market in recent years. At present, the common approach to programming gpu units is to write CPU specific code with low level gpu APIs such as CUDA. Although this approach can achieve good performance, it creates serious portability issues as programmers are required to write a specific version of the code for each potential target architecture. This results in high development and maintenance costs. We believe it is desirable to have a programming model which provides source code portability between CPUs and gpus, as well as different gpus. This would allow programmers to write one version of the code, which can be compiled and executed on either CPUs or gpus efficiently without modification. In this paper, we propose MapCG, a MapReduce framework to provide source code level portability between CPUs and gpus. In contrast to other approaches such as OpenCL, our framework, based on MapReduce, provides a high level programming model and makes programming much easier. We describe the design of MapCG, including the MapReduce-style high-level programming framework and the runtime system on the CPU and gpu. A prototype of the MapCG runtime, supporting multi-core CPUs and NVIDIA gpus, was implemented. Our experimental results show that this implementation can execute the same source code efficiently on multi-core CPU platforms and gpus, achieving an average speedup of 1.6-2.5x over previous implementations of MapReduce on eight commonly used applications.

关键词： portability parallel gpu programming

来源：评论

学校读者我要写书评

暂无评论

Optimal Path Maps on the gpu

引用

IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS 2020年第9期26卷 2863-2874页

作者： Farias, Renato Kallmann, Marcelo Univ Calif Comp Sci & Engn Dept Merced CA 95343 USA

We introduce a new method for computing optimal path maps on the gpu using OpenGL shaders. Our method explores gpu rasterization as a way to propagate optimal costs on a polygonal 2D environment, producing optimal path maps which can efficiently be queried at run-time. Our method is implemented entirely with gpu shaders, does not require pre-computation, addresses optimal path maps with multiple points and line segments as sources, and introduces a new optimal path map concept not addressed before: maps with weights at vertices representing possible changes in traversal speed. The produced maps offer new capabilities not explored by previous navigation representations and at the same time address paths with global optimality, a characteristic which has been mostly neglected in animated virtual environments. The proposed path maps partition the input environment into the regions sharing a same parent point along the shortest path to the closest source, taking into account possible speed changes at vertices. The proposed approach is particularly suitable for the animation of multiple agents moving toward the entrances or exits of a virtual environment, a situation which is efficiently represented with the proposed path maps.

关键词： Graphics processing units Path planning Microsoft Windows Geometry Navigation Virtual environments Computational efficiency Graphics data structures and data types path planning shortest path maps navigation gpu programming

来源：评论

学校读者我要写书评

暂无评论

Texture-based visualization of unsteady 3D flow by real-time advection and volumetric illumination

引用

IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS 2007年第3期13卷 569-582页

作者： Weiskopf, Daniel Schafhitzel, Tobias Ertl, Thomas Simon Fraser Univ Sch Comp Sci Graph Visualizat & Usabil Lab Burnaby BC V5A 1S6 Canada Univ Stuttgart Inst Visualizat & Interact Syst D-70569 Stuttgart Germany

This paper presents an interactive technique for the dense texture-based visualization of unsteady 3D flow, taking into account issues of computational efficiency and visual perception. High efficiency is achieved by a 3D graphics processing unit (gpu)-based texture advection mechanism that implements logical 3D grid structures by physical memory in the form of 2D textures. This approach results in fast read and write access to physical memory, independent of gpu architecture. Slice-based direct volume rendering is used for the final display. We investigate two alternative methods for the volumetric illumination of the result of texture advection: First, gradient-based illumination that employs a real-time computation of gradients, and, second, line-based lighting based on illumination in codimension 2. In addition to the Phong model, perception-guided rendering methods are considered, such as cool/warm shading, halo rendering, or color-based depth cueing. The problems of clutter and occlusion are addressed by supporting a volumetric importance function that enhances features of the flow and reduces visual complexity in less interesting regions. gpu implementation aspects, performance measurements, and a discussion of results are included to demonstrate our visualization approach.

关键词： scientific visualization time-dependent vector fields texture advection feature extraction volume visualization gpu programming

来源：评论

学校读者我要写书评

暂无评论

Automatic Quantization for Physics-Based Simulation

引用

ACM TRANSACTIONS ON GRAPHICS 2022年第4期41卷 p1-16页

作者： Liu, Jiafeng Shi, Haoyang Zhang, Siyuan Yang, Yin Ma, Chongyang Xu, Weiwei Zhejiang Univ State Key Lab CAD & CG Hangzhou Peoples R China Clemson Univ Clemson SC 29631 USA Univ Utah Salt Lake City UT 84112 USA Kuaishou Technol Beijing Peoples R China

Quantization has proven effective in high-resolution and large-scale simulations, which benefit from bit-level memory saving. However, identifying a quantization scheme that meets the requirement of both precision and memory efficiency requires trial and error. In this paper, we propose a novel framework to allow users to obtain a quantization scheme by simply specifying either an error bound or a memory compression rate. Based on the error propagation theory, our method takes advantage of auto-diff to estimate the contributions of each quantization operation to the total error. We formulate the task as a constrained optimization problem, which can be efficiently solved with analytical formulas derived for the linearized objective function. Our workflow extends the Taichi compiler and introduces dithering to improve the precision of quantized simulations. We demonstrate the generality and efficiency of our method via several challenging examples of physics-based simulation, which achieves up to 2.5x memory compression without noticeable degradation of visual quality in the results. Our code and data are available at https://***/Hanke98/AutoQuantizer.

关键词： Physics-based simulation Quantized computation gpu 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：