检索结果-南通市图书馆

Improving SIMD Utilization with Thread-Lane Shuffled Compaction in GPGPU

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2015年第4期24卷 684-688页

作者： LI Bingchao WEI Jizeng GUO Wei sun jizhou School of Computer Science and Technology Tianjin University

GPGPUs adopt SIMT execution model in which each logical thread in a warp corresponds to a SIMD lane while can still follow an independent control *** a branch divergence appears and threads within a warp take different execution paths, GPGPUs have to execute each path serially through SIMD lane masking, which potentially decreases the SIMD utilization and performance. We propose an efficient thread compaction mechanism to handle branch divergence with a novel register file structure. We also develop a new thread scheduling policy cooperating with our compaction mechanism. The simulation results show that our approach improves the SIMD utilization up to 74.4% and achieves a maximum 11.1%performance speedup with small hardware overhead.

关键词： Graphics processing unit(GPU) Singleinstruction multiple-data(SIMD) Branch divergence Register file

Bypass-Enabled Thread Compaction for Divergent Control Flow in Graphics Processing Units

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Shanghai Jiaotong university(Science) 2021年第2期26卷 245-256页

作者： LI Bingchao WEI Jizeng GUO Wei sun jizhou College of Computer Science and Technology Civil Aviation University of ChinaTianjin 300300China College of Intelligence and Computing Tianjin UniversityTianjin 300350China

Graphics processing units(GPUs)employ the single instruction multiple data(SIMD)hardware to run threads in parallel and allow each thread to maintain an arbitrary control *** running concurrently within a warp may jump to different paths after conditional *** divergent control flow makes some lanes idle and hence reduces the SIMD utilization of *** alleviate the waste of SIMD lanes,threads from multiple warps can be collected together to improve the SIMD lane utilization by compacting threads into idle ***,this mechanism induces extra barrier synchronizations since warps have to be stalled to wait for other warps for compactions,resulting in that no warps are scheduled in some *** this paper,we propose an approach to reduce the overhead of barrier synchronizat ions induced by compactions,In our approach,a compaction is bypassed by warps whose threads all jump to the same path after ***,warps waiting for a compaction can also bypass this compaction when no warps are ready for *** addition,a compaction is canceled if idle lanes can not be reduced via this *** experimental results demonstrate that our approach provides an average improvement of 21%over the baseline GPU for applications with massive divergent branches,while recovering the performance loss induced by compactions by 13%on average for applications with many non-divergent control flows.

关键词： graphics processing unit(GPU) single instruction ultiple data(SIMD) thread warps bypass

Reputation-based Trust Model in Grid Security System

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

通讯和计算机（中英文版） 2006年第8期3卷 41-46页

作者： Baolin Ma jizhou sun Ce Yu School of Electronic Information Engineering Tianfin University Tianjin 300072 China

Image-adaptive and Robust Digital Wavelet-domain Watermarking for Images

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

通讯和计算机（中英文版） 2005年第1期2卷 57-63页

作者： ZHANG Jianxun ZHANG Jiawan sun jizhou Electronic ＆ Information Engineering School Tianjin University. Tianjin 300072. China School of Computer Science ＆ Engineering Chongqing Institute of technology Chongqing 400050 China

Cleaning of Multi-Source Uncertain Time Series Data Based on PageRank

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Donghua University(English Edition) 2023年第6期40卷 695-700页

作者：高嘉伟孙纪舟 Faculty of Computer and Software Engineering Huaiyin Institute of TechnologyHuai’an 223003China

There are errors in multi-source uncertain time series *** discovery methods for time series data are effective in finding more accurate values,but some have limitations in their *** tackle this challenge,we propose a new and convenient truth discovery method to handle time series data.A more accurate sample is closer to the truth and,consequently,to other accurate *** the mutual-confirm relationship between sensors is very similar to the mutual-quote relationship between web pages,we evaluate sensor reliability based on PageRank and then estimate the truth by sensor ***,this method does not rely on smoothness assumptions or prior knowledge of the ***,we validate the effectiveness and efficiency of the proposed method on real-world and synthetic data sets,respectively.

关键词： big data data cleaning time series truth discovery PageRank

Real-Time Arm Animation with Natural Collision Avoidance

在线全文

学校读者我要写书评

暂无评论

Transactions of Tianjin University 2007年第5期13卷 365-369页

作者：倪慧孙济洲刘艳 School of Computer Science and Technology Tianjin University

A novel approach to realistic collision-free animation of the upper limb was *** to the obstacle-avoidance strategy of human hand,the movement trajectory was computed by manipulability model and minimum-jerk based *** each key frame,an improved in～verse kinematics method was adopted to obtain a believable posture of the upper *** comparing with the real movement data obtained from the motion capture device,the resultant animation was testified to be *** method can be performed in interactive time,and therefore is applicable in animation edit and control of virtual humans.

关键词： animation collision avoidance manipulability model interpolation

在线全文

学校读者我要写书评

暂无评论

基于系统调用顺序和频度特性的入侵检测模型

计算机工程 2006年第13期32卷 18-19,43页

作者：张桂玲孙济洲天津大学电信学院天津300072

提出了一种将系统调用的顺序特性和频度特性相接合来构建入侵检测模型(COFIDS模型)的新方法,该模型采用kNN(k-NearestNeighbor Classifier)算法实现入侵检测,并利用一种改进的相似因子,来增加系统调用序列间相似度的差别,减少了识别误差... 详细信息

提出了一种将系统调用的顺序特性和频度特性相接合来构建入侵检测模型(COFIDS模型)的新方法,该模型采用kNN(k-NearestNeighbor Classifier)算法实现入侵检测,并利用一种改进的相似因子,来增加系统调用序列间相似度的差别,减少了识别误差,提高了检测率,降低了入侵检测的误报率。实验表明,COFIDS还具有较强的抗噪声干扰的能力。

关键词：入侵检测系统调用 kNN算法

在线全文

学校读者我要写书评

暂无评论

基于改进的动态神经网络的网络入侵检测模型

计算机工程 2006年第11期32卷 10-12页

作者：张桂玲孙济洲天津大学电信学院计算机软件研究与发展中心天津300072

提出了一种改进型的动态神经网络,并成功地将其应用于网络入侵检测系统中。对于给定的全连接的动态神经网络,在通过学习以后可以成为部分连接的神经网络系统,从而降低了计算的成本。针对目前常见的4种不同类型的网络攻击行为(即DoS,Prob... 详细信息

提出了一种改进型的动态神经网络,并成功地将其应用于网络入侵检测系统中。对于给定的全连接的动态神经网络,在通过学习以后可以成为部分连接的神经网络系统,从而降低了计算的成本。针对目前常见的4种不同类型的网络攻击行为(即DoS,Probe,R2L,和U2R),利用给定的改进型的动态神经网络分别构建相对应的检测系统。然后使用改进的遗传算法对给定的动态神经网络的权值和开关参数进行调节,以适应不同类型的入侵检测。最后利用KDD’99网络入侵检测数据对所提出的网络入侵检测模型进行训练和测试,初步试验结果表明,所提出的入侵检测系统具有较高的检测率。

关键词：动态神经网络网络入侵检测遗传算法

在线全文

学校读者我要写书评

暂无评论

基于像素光照计算的硬件加速技术

计算机工程 2005年第18期31卷 34-35,104页

作者：王洁宁孙济洲天津大学电信学院计算机科学系

给出了基于像素光照计算的原理和硬件体系结构;提出了一种基于像素光照计算的设计分析方法及相应的光照计算参数的纹理表示模型;通过凹凸纹理的实时绘制,给出了不同硬件加速模式下的实现结果。

关键词：实时绘制硬件加速凹凸纹理 Phong光照模型 Cg