检索结果-内蒙古大学图书馆

您好，读者！请登录

内蒙古大学图书馆

首页
概况
党建
资源
服务
科研支持
- 论文收录引用证明
- 科技查新
知识产权
档案馆
帮助

咨询与建议

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

您的常用邮箱：*

您的手机号码：*

问题描述：

当前已输入0个字，您还可以输入200个字

全部搜索
期刊论文
图书
学位论文
标准
纸本馆藏
外文资源发现
数据库导航
超星发现

高级检索

时间限定

出版年份：

文献类型

图书期刊文献学位论文多媒体

馆藏选择

电子馆藏纸本馆藏

核心期刊

全部期刊 SCI 收录期刊 SSCI 收录期刊 EI 收录期刊 CSCD 收录期刊 CSSCI 收录期刊

语言

中文英文

文献类型

期刊文献图书学位论文标准纸本馆藏

帮助

文字说明：

T=题名（书名、题名），A=作者（责任者），K=主题词，P=出版物名称，PU=出版社名称，O=机构（作者单位、学位授予单位、专利申请人），L=中图分类号，C=学科分类号，U=全部字段，Y=年（出版发行年、学位年度、标准发布年）

检索规则说明：

AND代表“并且”；OR代表“或者”；NOT代表“不包含”；(注意必须大写,运算符两边需空一格)

检索范例：

范例一：(K=图书馆学 OR K=情报学) AND A=范并思 AND Y=1982-2016
范例二：P=计算机应用与软件 AND (U=C++ OR U=Basic) NOT K=Visual AND Y=2011-2016

分类表

所选分类

>> <<

限定检索结果

文献类型

33,189 篇 会议
834 篇 期刊文献
231 册 图书
1 篇 学位论文
1 篇 科技报告

馆藏范围

34,256 篇 电子文献
0 种 纸本馆藏

日期分布

学科分类号

20,166 篇 工学
- 17,877 篇 计算机科学与技术...
- 7,815 篇 软件工程
- 4,550 篇 电气工程
- 2,846 篇 信息与通信工程
- 1,465 篇 控制科学与工程
- 748 篇 电子科学与技术（可...
- 690 篇 动力工程及工程热...
- 473 篇 机械工程
- 381 篇 生物工程
- 339 篇 石油与天然气工程
- 303 篇 仪器科学与技术
- 248 篇 生物医学工程（可授...
- 248 篇 网络空间安全
- 247 篇 环境科学与工程（可...
- 217 篇 力学（可授工学、理...
- 206 篇 建筑学
- 200 篇 材料科学与工程（可...
- 191 篇 安全科学与工程
- 186 篇 土木工程
4,383 篇 理学
- 2,842 篇 数学
- 888 篇 物理学
- 532 篇 系统科学
- 471 篇 生物学
- 424 篇 统计学（可授理学、...
- 230 篇 化学
3,206 篇 管理学
- 2,476 篇 管理科学与工程(可...
- 1,264 篇 工商管理
- 965 篇 图书情报与档案管...
287 篇 经济学
- 281 篇 应用经济学
287 篇 医学
- 217 篇 临床医学
191 篇 法学
98 篇 教育学
98 篇 农学
77 篇 文学
10 篇 艺术学
9 篇 军事学
1 篇 哲学

主题

4,290 篇 grid computing
4,221 篇 distributed comp...
2,063 篇 concurrent compu...
2,046 篇 cloud computing
2,017 篇 parallel process...
1,844 篇 computational mo...
1,619 篇 computer archite...
1,403 篇 computer science
1,362 篇 application soft...
1,288 篇 resource managem...
1,257 篇 computer network...
988 篇 processor schedu...
781 篇 scalability
776 篇 distributed comp...
765 篇 high performance...
760 篇 peer to peer com...
682 篇 costs
670 篇 distributed data...
666 篇 algorithm design...
651 篇 middleware

机构

51 篇 department of co...
47 篇 school of comput...
46 篇 institute of com...
45 篇 college of compu...
39 篇 huazhong univ sc...
35 篇 china electric p...
34 篇 shanghai jiao to...
33 篇 school of comput...
32 篇 school of comput...
31 篇 univ chinese aca...
31 篇 department of co...
27 篇 inria
27 篇 tsinghua univ de...
26 篇 university of ch...
25 篇 shandong provinc...
24 篇 tsinghua univers...
23 篇 huazhong univers...
23 篇 institute of par...
23 篇 nanjing univ sta...
23 篇 college of intel...

作者

69 篇 jin hai
51 篇 hai jin
49 篇 rajkumar buyya
39 篇 fahringer thomas
37 篇 zomaya albert y.
33 篇 liu yang
31 篇 li dongsheng
29 篇 dongsheng li
29 篇 m. takizawa
27 篇 duerr frank
26 篇 li li
24 篇 xhafa fatos
23 篇 wei li
22 篇 wang lei
22 篇 yang yang
20 篇 lei wang
20 篇 xuejun yang
19 篇 dongarra jack
18 篇 singh bhim
18 篇 bala piotr

语言

33,705 篇 英文
372 篇 其他
145 篇 中文
43 篇 俄文
2 篇 乌克兰文
1 篇 德文
1 篇 西班牙文
1 篇 法文
1 篇 葡萄牙文
1 篇 土耳其文

检索条件"任意字段=International Conference on Parallel Distributed and Grid Computing"

共 34256 条记录，以下是371-380 订阅

全选清除本页清除全部题录导出标记到"检索档案"

详细简洁

排序：

Research on Heterogeneous computing Technology Based on Domestic CPUs and Reconfigurable Chips 4

Research on Heterogeneous Computing Technology Based on Dome...

引用

4th international conference on Intelligent Power and Systems, ICIPS 2024

作者： Guan, Xiaojuan Wang, Chen Zhou, Cheng Zhu, Sheng China Electric Power Research Institute Co. Ltd State Grid Laboratory of Power Cyber-Security Protection and Monitoring Technology Nanjing China

ISBN: (纸本)9798331532765

In view of the performance and security problems existing after the replacement of specialized security protection equipment for power grids with domestic software and hardware, and in response to the security protection requirements of different application scenarios, the research on heterogeneous fusion computing technology based on domestic CPUs and reconfigurable chips is proposed. The optimal mapping of national secret algorithms, isolation exchange, protocol processing, and data analysis on reconfigurable chips is realized, supporting the parallel computing of special processors for scene acceleration and domestic CPUs, and meeting the requirements of different network security analysis and computing scenarios in the power system. © 2024 IEEE.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

A General parallel Framework for Material Point Method Based on the p4est Library 22

A General Parallel Framework for Material Point Method Based...

引用

22nd IEEE international Symposium on parallel and distributed Processing with Applications, ISPA 2024

作者： Tian, Shaobo Wang, Jiaheng Li, Jialin Ye, Huang Zhang, Jian Chinese Academy of Sciences Computer Network Information Center China University of Chinese Academy of Sciences China Renmin University of China China

ISBN: (纸本)9798331509712

Material Point Method(MPM) is widely used to simulate large deformation processes such as material fracture, collision, and fluid structure interaction. In general, an MPM simulation evolves a dynamic process in which a large number of Lagrangian particles move on top of Eulerian grids. During the process, physical quantities such as momentum and force are interpolated frequently between the particles and their corresponding grids. High Performance computing(HPC) has become an indispensable tool for large-scale MPM simulations, while dynamic task decomposition and load balancing are critical for *** this paper, we present a general parallel framework integrating MPM and a well-established oct-tree mesh management library, p4est, aiming to improve the efficiency of large-scale MPM simulations. Through careful design of data structure and interfaces to connect the original grid class of MPM with the p4est mesh, a highly modular structure is realized for the framework. This design allows the users to easily add new features or optimize existing ones, thereby enhancing its flexibility and reusability. On top of this, dynamic load-balancing strategies for MPM can be realized without much effort. A recommended strategy that considers both particle and grid workload is presented. In addition, the advantage of dynamic load balancing is demonstrated and analyzed through practical simulations. The versatility and efficiency of the proposed framework are also demonstrated through concrete real-world applications including penetration and building implosion simulations with up to 1.5 billion degrees of freedom. The code achieves 87% overall parallel efficiency scaling up to 2048 processors. © 2024 IEEE.

关键词： Mesh generation

来源：评论

学校读者我要写书评

暂无评论

Accelerating Large Language Model Training with Hybrid GPU-based Compression 24

Accelerating Large Language Model Training with Hybrid GPU-b...

引用

24th IEEE/ACM international Symposium on Cluster, Cloud, and Internet computing (CCgrid)

作者： Xu, Lang Anthony, Quentin Zhou, Qinghua Alnaasan, Nawras Gulhane, Radha Shafi, Aamir Subramoni, Hari Panda, Dhabaleswar K. (Dk) Ohio State Univ Columbus OH 43210 USA

ISBN: (纸本)9798350395679;9798350395662

Data parallelism (DP), Tensor parallelism (TP), and Pipeline parallelism (PP) are the three strategies widely adopted to enable fast and efficient Large Language Model (LLM) training. However, these approaches rely on data-intensive communication routines to collect, aggregate, and re-distribute gradients, activations, and other important model information, which pose significant overhead. Co-designed with GPU-based compression libraries, MPI libraries have been proven to reduce message size significantly, and leverage interconnect bandwidth, thus increasing training efficiency while maintaining acceptable accuracy. In this work, we investigate the efficacy of compression-assisted MPI collectives under the context of distributed LLM training using 3D parallelism and ZeRO optimizations. We scaled up to 192 V100 GPUs on the Lassen supercomputer. First, we enabled a naive compression scheme across all collectives and observed a 22.5% increase in TFLOPS per GPU and a 23.6% increase in samples per second for GPT-NeoX-20B training. Nonetheless, such a strategy ignores the sparsity discrepancy among messages communicated in each parallelism degree, thus introducing more errors and causing degradation in training loss. Therefore, we incorporated hybrid compression settings toward each parallel dimension and adjusted the compression intensity accordingly. Given their low-rank structure [1], we apply aggressive compression on gradients when performing DP All-reduce. We adopt milder compression to preserve precision while communicating activations, optimizer states, and model parameters in TP and PP. Using the adjusted hybrid compression scheme, we demonstrate a 17.3% increase in TFLOPS per GPU and a 12.7% increase in samples per second while reaching baseline loss convergence. *

关键词： All-reduce Large Language Model Compression GPU-Aware MPI Deep Learning distributed Training

来源：评论

学校读者我要写书评

暂无评论

Research and Application of Three-Dimensional Point Cloud Data Analysis Technology for Smart grid Transmission Data Based on Cloud computing 3rd

Research and Application of Three-Dimensional Point Cloud Da...

引用

3rd international Symposium on New Energy and Electrical Technology (ISNEET)

作者： Guo, Jinchao Rao, Chengcheng Chen, Yun Guangdong Power Grid Co Ltd Machine Patrol Management Ctr Guangzhou 521000 Peoples R China

ISBN: (纸本)9789819905553;9789819905539;9789819905522

Big data is an important product of the information age. Integrating big data into smart grid applications and correctly grasping the key technologies of big data can effectively promote the sustainable development of power industry and the construction of strong smart grid. As far as modern smart grid is concerned, this is both an opportunity and a challenge. 3D point cloud data processing is the core content of reverse engineering technology. As an important data processing step in the preprocessing stage of 3D point cloud, point cloud registration plays an extremely important role in obtaining the complete 3Dcoordinates of the measured target surface. However, at present, the registration speed, accuracy and reliability of various registration algorithms still need to be improved. Cloud computing technology integrates several cheap ordinary PCs into a cloud computing cluster, which realizes the safe storage and efficient processing of massive data. Therefore, consider combining cloud computing with data mining algorithm to solve the problem of massive data conversion in smart grid. In this paper, cloud computing technology is introduced into the smart grid condition monitoring field. By introducing distributed file system, improving traditional density clustering algorithm and parallel design, the storage and clustering division of big data in condition monitoring are effectively solved, which provides a feasible method for the application of cloud computing in condition monitoring field.

关键词： Cloud computing Smart grid Transmission data 3D point cloud data analysis technology

来源：评论

学校读者我要写书评

暂无评论

EBFT: Efficient BFT Consensus through Shortcut Replies 21

EBFT: Efficient BFT Consensus through Shortcut Replies

引用

21st IEEE international Symposium on parallel and distributed Processing with Applications, 13th IEEE international conference on Big Data and Cloud computing, 16th IEEE international conference on Social computing and Networking and 13th international conference on Sustainable computing and Communications, ISPA/BDCloud/SocialCom/SustainCom 2023

作者： Lei, Qingsheng Xia, Yongjun Guo, Qitao Hao, Rui Dai, Xiaohai Dai, Weiqi Economic & Technology Research Institute State Grid Hubei Electric Power Company Limited China Hubei Central China Technology Development of Electric Power Company Limited China Software Institute Nanjing University China Huazhong University of Science and Technology School of Computer Science and Technology China Huazhong University of Science and Technology School of Cyber Science and Engineering China

ISBN: (纸本)9798350329223

As the core building block of blockchain technology, Byzantine Fault Tolerance (BFT) consensus has gathered renewed attention during the past decade. A large number of BFT consensus protocols have been proposed since its inception in the 1980s, the best-known of which is Practical Byzantine Fault Tolerance (PBFT). Generally speaking, PBFT consists of two sub-protocols: normal-case protocol and view-change protocol. However, it requires 7 δ for a client to commit a request in the normal case, where δ represents the actual network delay. If a request happens to go through the view change process, it will take additional 7 δ for a client to commit a request. To reduce the consensus latency, we propose a novel BFT consensus protocol named EBFT, which introduces the shortcut replies to the consensus design. With these shortcut replies, EBFT reduces the normal-case latency and view-change latency to 5 δ and 2 δ, respectively. Extensive experiments are conducted to evaluate EBFT, whose results demonstrate its effectiveness. © 2023 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

parallelized Multi-Agent Bayesian Optimization in Lava

Parallelized Multi-Agent Bayesian Optimization in Lava

引用

2024 international conference on Neuromorphic Systems, ICONS 2024

作者： Snyder, Shay Gobin, Derek Clerico, Victoria Risbud, Sumedh R. Parsa, Maryam George Mason University Electrical and Computer Engineering Fairfax United States Intel Labs Neuromorphic Computing Lab Santa Clara United States

ISBN: (纸本)9798350368659

In parallel with the continuously increasing parameter space dimensionality, search and optimization algorithms should support distributed parameter evaluations to reduce cumulative runtime. Intel's neuromorphic optimization library, Lava-Optimization, was introduced as an abstract optimization system compatible with neuromorphic systems developed in the broader Lava software framework. In this work, we introduce Lava Multi-Agent Optimization (LMAO) with native support for distributed parameter evaluations communicating with a central Bayesian optimization system. LMAO provides an abstract framework for deploying distributed optimization and search algorithms within the Lava software framework. Moreover, LMAO introduces support for random and grid search along with process connections across multiple levels of mathematical precision. We evaluate the algorithmic performance of LMAO with a traditional non-convex optimization problem, a fixed-precision transductive spiking graph neural network for citation graph classification, and a neuromorphic satellite scheduling problem. Our results highlight LMAO's efficient scaling to multiple processes, reducing cumulative runtime and minimizing the likelihood of converging to local optima. © 2024 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Research on the Economic Scheduling of Multi-distributed Power Sources Based on the Super-Node Cooperative ADMM Algorithm 5

Research on the Economic Scheduling of Multi-Distributed Pow...

引用

5th international conference on Artificial Intelligence Technologies and Applications, ICAITA 2023

作者： Li, Qiang Ge, Zhiping Zhang, Ruishan Wang, Changyi Zhang, Yanbin Lanzhou Longneng Power Technology Co. Gansu Province Lanzhou China

ISBN: (纸本)9781643684840

To address the problems of large computation and communication in the economic scheduling of multi-distributed power supplies and the difficulty of overall optimization, an economic scheduling method of distributed power supplies based on the supernode cooperative ADMM algorithm is proposed. First, the supernode selection and local set partitioning are performed on the basis of the original communication topology diagram of the distributed power network, with supernodes solving decision variables in parallel and ordinary nodes in the local set only communicating with supernodes;second, the standard ADMM algorithm is improved so that each node searches for the optimal solution of its subproblem in parallel;in a distributed power grid containing multiple renewable energy sources, the operation cost minimization optimization The proposed method is used to solve the large-scale optimization model in a distributed grid with multiple renewable energy sources. Finally, the proposed algorithm is verified to accelerate the convergence speed of iterative optimization search through simulation cases, which can effectively improve the convergence speed of distributed power cooperative consistency. © 2024 The Authors.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

A Survey of Edge computing Resource Allocation and Task Scheduling Optimization 5th

A Survey of Edge Computing Resource Allocation and Task Sche...

引用

5th international conference on Big Data and Security (ICBDS)

作者： Xu, Xiaowei Ding, Han Wang, Jiayu Hua, Liang State Grid Wuxi Power Supply Co Wuxi Jiangsu Peoples R China

ISBN: (纸本)9789819743896;9789819743902

The optimization problem of resource allocation and task scheduling involves effectively utilizing limited computing resources and arranging user service requests or tasks in a distributed computing environment to achieve improved performance, reduced costs, and energy savings. With the advancements in technologies such as the Internet of Things and 5G, edge computing has emerged as a new paradigm that deploys data processing and analysis capabilities closer to the data source, enabling faster, more secure, and reliable services. Consequently, the widespread adoption of edge computing presents new challenges and opportunities for resource allocation and task scheduling optimization, including the management of massive data streams, adaptation to dynamic network environments, and coordination between edge and cloud computing. This paper summarizes the issues, evaluation dimensions and methods of resource allocation and task scheduling optimization in edge computing and identifies future prospects and challenges in this area.

关键词： edge computing Internet of Things resource allocation task scheduling

来源：评论

学校读者我要写书评

暂无评论

Balancing Power in a grid-Linked Inverter with an Unbalanced grid Through parallel Current Control

Balancing Power in a Grid-Linked Inverter with an Unbalanced...

引用

2024 international conference on Innovation and Intelligence for Informatics, computing, and Technologies, 3ICT 2024

作者： Soth, Panha Choeung, Chivon San, Sokna Yay, Socheat Tang, Heng Torn, Vanna In, Sokvan Cheng, Horchhong Dept. of Control System National Polytechnic Institute of Cambodia Phnom Penh Cambodia Dept. of Control System NPIC Phnom Penh Cambodia Dept. of Electrical Installation NPIC Phnom Penh Cambodia Dept. of Gen. Transmisssion and Sub-Transmission EAC Phnom Penh Cambodia Faculty of Electricity NPIC Phnom Penh Cambodia

ISBN: (纸本)9798331533137

This paper presents an approach for power balancing in grid-linked inverters within the context of smart grids, addressing the challenge of unbalanced grid voltage conditions. The proposed method employs a paralleled current control strategy, utilizing state feedback integral control to regulate both positive and negative sequence currents. This approach ensures resilient performance under grid voltage imbalances, which are a common problem in smart grid environments. The control gains are optimized through a linear matrix inequality-based method, enhancing the system’s stability and responsiveness to fluctuating grid conditions. Simulations conducted using MATLAB and PSIM demonstrate the effectiveness of this control strategy in maintaining power balance and achieving a unity power factor despite unbalanced grid voltages. This study aims to contribute to the improvement of smart grid technologies by offering a dependable and efficient solution for power balancing in grid-connected inverter systems. ©2024 IEEE.

关键词： Smart power grids

来源：评论

学校读者我要写书评

暂无评论

Towards Energy-Aware Machine Learning in Geo-distributed IoT Settings

Towards Energy-Aware Machine Learning in Geo-Distributed IoT...

引用

29th international conference on parallel and distributed computing (Euro-Par)

作者： Trihinas, Demetris Thamsen, Lauritz Univ Nicosia Dept Comp Sci Nicosia Cyprus Univ Glasgow Sch Comp Sci Glasgow Lanark Scotland

ISBN: (纸本)9783031488023;9783031488030

As the Internet of Things (IoT) increasingly empowers the network extremes with in-place intelligence through Machine Learning (ML), energy consumption and carbon emissions become crucial factors. ML is often computationally intensive, with state-of-the-art model architectures consuming significant energy per training round and imposing a large carbon footprint. This work, therefore, argues for the need to introduce novel mechanisms into the ML pipelines of IoT services, so that energy awareness is integrated in the decision-making process for when and where to initiate ML model training.

关键词： Machine Learning Internet of Things distributed Systems Energy Profiling Carbon Footprint System Orchestration

来源：评论

学校读者我要写书评

暂无评论

没有更多数据了...

全选清除本页清除全部题录导出标记到“检索档案”

共500页 << < 34 35 36 37 38 39 40 41 42 43 > >>

检索报告对象比较合并检索0

隐藏清空

合并搜索

回到顶部

执行限定条件

内容：

评分：

请选择保存的检索档案：

请选择收藏分类：

订阅名称：

通借通还

温馨提示：

图书名称：

借书校区：

取书校区：

手机号码：

邮箱地址：

一卡通帐号：

电话和邮箱必须正确填写，我们会与您联系确认。

联系人：

所在院系：

联系邮箱：

联系电话：

内蒙古自治区呼和浩特市赛罕区大学西街235号邮编: 010021

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：