检索结果-内蒙古大学图书馆

Full Eulerian deformable solid-fluid interaction scheme based on building-cube method for large-scale parallel computing

引用

INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING 2019年第2期117卷 221-248页

作者： Nishiguchi, Koji Bale, Rahul Okazawa, Shigenobu Tsubokura, Makoto RIKEN RIKEN Ctr Computat Sci Kobe Hyogo Japan Univ Yamanashi Dept Mech Engn Fac Engn Kofu Yamanashi Japan Kobe Univ Grad Sch Syst Informat Dept Computat Sci Kobe Hyogo Japan

We propose a full Eulerian incompressible solid-fluid interaction scheme capable of achieving high parallel efficiency and easily generating meshes for complex solid geometries. While good scalability of a full Eulerian solid-fluid interaction formulation has been reported by Sugiyama et al, their analysis was carried out using uniform Cartesian mesh and an artificial compressibility method. Typically, it is more challenging to achieve good scalability for hierarchical Cartesian meshes and a fully incompressible formulation. In addition, the conventional full Eulerian methods require a large computational cost to resolve complex solid geometries due to the usage of uniform Cartesian meshes. In an attempt to overcome the aforementioned issues, we employ the building-cube method, where the computational domain is divided into cubic regions called cubes. Each cube is divided at equal intervals, the same number of cubes is assigned to each core, and the spatial loop processing is executed for each cube. The numerical method is verified by computing five numerical examples. In the weak scaling test, the parallel efficiency at 32768 cores with 32 cores as a reference is 93.6%. In the strong scaling test, the parallel efficiency at 32768 cores with 128 cores as a reference is 70.2%.

关键词： building-cube method Eulerian formulation fluid-structure interaction hyperelasticity large deformation large-scale parallel computing

来源：评论

学校读者我要写书评

暂无评论

A heterogeneous accelerated simulation framework for wind field dynamic model

引用

IET RENEWABLE POWER GENERATION 2023年第1期17卷 137-149页

作者： Li, Bing Zhao, Haoran Zhang, Kunpeng Shandong Univ Sch Elect Engn Jinan 250061 Peoples R China

This paper proposes an efficient and extensible simulation framework for mid-fidelity wind field model based on heterogeneous architecture. The wind field model consists of 3-D wind inflow, wake dynamics, and wind turbine models. The CPU-GPU heterogeneous platform accelerates the 3-D wind inflow model simulation by performing large-scale parallel computing based on GPU. The CPU is utilized to dispatch the simulation and perform serial computing. Compared with the conventional simulation tool, the simulation speed of the 3-D wind inflow model is highly improved, which fulfils the requirement of the real-time wind field simulation. The hardware architecture of the framework is designed to be highly extensible by multiple computing nodes for 3-D wind inflow, wake dynamic, and wind turbine model. The extendability enables the ability for large-scale wind field simulation with high efficiency. A 5 km$\rm km$ x 2 km$\rm km$ x 0.35 km$\rm km$ wind field consisting of three wind turbines is used as the case. The case study was conducted and analyzed, proving that the framework can capture the detailed wake dynamic, wind turbine dynamic, and wind inflow dynamic. Meanwhile, the calculation speed of disturbed wind speed increased by 264 times.

关键词： heterogeneous architecture wind simulation speed Wind power plants wakes wind field dynamic model real-time wind field simulation heterogeneous accelerated simulation framework large-scale wind field simulation wind turbines CPU-GPU heterogeneous platform extensible simulation framework wind turbine dynamic disturbed wind speed large-scale parallel computing Power engineering computing power engineering computing parallel software wind inflow dynamic mid-fidelity wind field model wake dynamic efficient simulation framework graphics processing units wind turbine model

来源：评论

学校读者我要写书评

暂无评论

A high fidelity general purpose 3-D Monte Carlo particle transport program JMCT3.0

引用

Nuclear Science and Techniques 2022年第8期33卷 175-192页

作者： Li Deng Gang Li Bao-Yin Zhang Rui Li Ling-Yu Zhang Xin Wang Yuan-Gang Fu Dun-Fu Shi Peng Liu Yan Ma Dan-Hu Shangguan Ze-Hua Hu Sheng-Cheng Zhou Jing-Wen Shen Institute of Applied Physics and Computational Mathematics(IAPCM) Beijing 100094China CAEP Software Centre for High Performance Numerical Simulation(CAEP-SCNS) Beijing 100088China

JMCT is a large-scale,high-fidelity,three-dimensional general neutron–photon–electron–proton transport Monte Carlo software *** was developed based on the combinatorial geometry parallel infrastructure JCOGIN and the adaptive structured mesh infrastructure *** is equipped with CAD modeling and visualizes the image *** supports the geometry of the body and the structured/unstructured *** has most functions,variance reduction techniques,and tallies of the traditional Monte Carlo particle transport *** energy models,multi-group and continuous,are *** recent years,some new functions and algorithms have been developed,such as Doppler broadening on-thefly(OTF),uniform tally density(UTD),consistent adjoint driven importance sampling(CADIS),fast criticality search of boron concentration(FCSBC)domain decomposition(DD),adaptive control rod moving(ACRM),and random geometry(RG)*** JMCT is also coupled with the discrete ordinate SNcode JSNT to generate source-biasing factors and weight-window *** present,the number of geometric bodies,materials,tallies,depletion zones,and parallel processors are sufficiently large to simulate extremely complicated device *** can be used to simulate reactor physics,criticality safety analysis,radiation shielding,detector response,nuclear well logging,and dosimetry calculations *** particular,JMCT can be coupled with depletion and thermal-hydraulics for the simulation of reactor nuclear-hot feedback *** paper describes the progress in advanced modeling,high-performance numerical simulation of particle transport,multiphysics coupled calculations,and large-scale parallel computing.

关键词： Advanced modeling High-performance numerical simulation Multi-physics coupled calculation large-scale parallel computing JMCT

来源：评论

学校读者我要写书评

暂无评论

MLPs: Efficient Training of MiniGo on large-scale Heterogeneous computing System 28

MLPs: Efficient Training of MiniGo on Large-scale Heterogene...

引用

IEEE 28th International Conference on parallel and Distributed Systems (IEEE ICPADS)

作者： Qiao, Peng He, Zhouyu Li, Rongchun Jiang, Jingfei Dou, Yong Li, Dongsheng Natl Univ Def Technol Sci & Technol Parallel & Distributed Lab Changsha Peoples R China

ISBN: (纸本)9781665473156

Deep Reinforcement Learning has been successfully applied in various applications and achieved impressive performance compared with previous traditional methods but suffers from high computation cost and long training time. MLPerf takes deep reinforcement learning as one of the benchmark tracks and provides a single node training version of MiniGo as a reference. A key challenge is to achieve efficient MiniGo training on a large-scale computing system. According to the training computation pattern in MiniGo and the characteristics of our large-scale heterogeneous computing system, we propose a MultiLevel parallel strategy, MLPs, including task-level parallelism between nodes, CPU-DSP heterogeneous parallelism, and DSP multi-core parallelism. The proposed method reduces the overall execution time from 43 hours to 16 hours while scaling the node size from 1067 to 4139. The scaling efficiency is 69.1%. According to our fitting method, the scaling efficiency is 46.5% when scaling to 8235 nodes. The experimental results show that the proposed method achieves the efficient training of MiniGo on the largescale heterogeneous computing system.

关键词： deep reinforcement learning deep neural networks MLPerf heterogeneous architecture large-scale parallel computing

来源：评论

学校读者我要写书评

暂无评论

Survey on memory management techniques in heterogeneous computing systems

引用

IET COMPUTERS AND DIGITAL TECHNIQUES 2020年第2期14卷 47-60页

作者： Hazarika, Anakhi Poddar, Soumyajit Rahaman, Hafizur Indian Inst Informat Technol Guwahati Gauhati India Indian Inst Engn Sci & Technol Sibpur India

A major issue faced by data scientists today is how to scale up their processing infrastructure to meet the challenge of big data and high-performance computing (HPC) workloads. With today's HPC domain, it is required to connect multiple graphics processing units (GPUs) to accomplish large-scale parallel computing along with CPUs. Data movement between the processor and on-chip or off-chip memory creates a major bottleneck in overall system performance. The CPU/GPU processes all the data on a computer's memory and hence the speed of the data movement to/from memory and the size of the memory affect computer speed. During memory access by any processing element, the memory management unit (MMU) controls the data flow of the computer's main memory and impacts the system performance and power. Change in dynamic random access memory (DRAM) architecture, integration of memory-centric hardware accelerator in the heterogeneous system and Processing-in-Memory (PIM) are the techniques adopted from all the available shared resource management techniques to maximise the system throughput. This survey study presents an analysis of various DRAM designs and their performances. The authors also focus on the architecture, functionality, and performance of different hardware accelerators and PIM systems to reduce memory access time. Some insights and potential directions toward enhancements to existing techniques are also discussed. The requirement of fast, reconfigurable, self-adaptive memory management schemes in the high-speed processing scenario motivates us to track the trend. An effective MMU handles memory protection, cache control and bus arbitration associated with the processors.

关键词： multiprocessing systems parallel processing cache storage computer architecture graphics processing units DRAM chips large-scale parallel computing off-chip memory memory management unit data flow dynamic random access memory architecture memory-centric hardware accelerator heterogeneous system processing-in-memory PIM systems memory access time self-adaptive memory management schemes high-speed processing scenario memory protection memory management techniques heterogeneous computing systems data scientists today processing infrastructure big data high-performance computing workloads HPC workloads MMU controls resource management techniques DRAM architecture multiple graphics processing units CPU GPU processes cache control

来源：评论

学校读者我要写书评

暂无评论

Simulation of Challenging Electromagnetic Problems Using a Massively parallel Finite Element Method Solver

引用

IEEE ACCESS 2019年 7卷 20346-20362页

作者： Zuo, Sheng Donoro, Daniel Garcia Zhang, Yu Bai, Yang Zhao, Xunwang Xidian Univ Shaanxi Key Lab Large Scale Electromagnet Comp Xian 710071 Shaanxi Peoples R China

This communication presents an efficient massively parallel finite element method solver for the solution of complex and electrically large electromagnetic problems with arbitrary structures. The solver makes use of a domain decomposition algorithm to decompose the original problem into several non-overlapping sub-domains that may be solved independently in parallel through the application of the corresponding transmission conditions on the interfaces of the adjacent sub-domains. A numerical exact mesh truncation algorithm called finite element-iterative integral equation evaluation, accelerated with multilevel fast multipole algorithm, is implemented to meet the highly accurate requirements of today's challenging simulations. What's more, a hybrid message passing interface and an open multi-processing parallel framework are designed to achieve large-scale parallel performance on supercomputers. Through several numerical examples, the accuracy, effectiveness, and scalability of the proposed solver will be demonstrated, achieving more than 60% parallel efficiency on an eight times CPU core scale (from 1280 to 10240 cores).

关键词： Finite element method (FEM) domain decomposition method (DDM) large-scale parallel computing finite element-iterative integral equation evaluation (FE-IIEE) ten thousand CPU cores

来源：评论

学校读者我要写书评

暂无评论

Rateless Codes for Near-Perfect Load Balancing in Distributed Matrix-Vector Multiplication

引用

PROCEEDINGS OF THE ACM ON MEASUREMENT AND ANALYSIS OF computing SYSTEMS 2019年第3期3卷 1–40页

作者： Mallick, Ankur Chaudhari, Malhar Sheth, Utsav Palanikumar, Ganesh Joshi, Gauri Carnegie Mellon Univ Pittsburgh PA 15213 USA Oracle Corp Redwood City CA USA Automat Anywhere San Jose CA USA Apple Inc Cupertino CA 95014 USA

large-scale machine learning and data mining applications require computer systems to perform massive matrix-vector and matrix-matrix multiplication operations that need to be parallelized across multiple nodes. The presence of straggling nodes - computing nodes that unpredictably slowdown or fail - is a major bottleneck in such distributed computations. Ideal load balancing strategies that dynamically allocate more tasks to faster nodes require knowledge or monitoring of node speeds as well as the ability to quickly move data. Recently proposed fixed-rate erasure coding strategies can handle unpredictable node slowdown, but they ignore partial work done by straggling nodes thus resulting in a lot of redundant computation. We propose a rateless fountain coding strategy that achieves the best of both worlds - we prove that its latency is asymptotically equal to ideal load balancing, and it performs asymptotically zero redundant computations. Our idea is to create linear combinations of them rows of the matrix and assign these encoded rows to different worker nodes. The original matrix-vector product can be decoded as soon as slightly more than m row-vector products are collectively finished by the nodes. We conduct experiments in three computing environments: local parallel computing, Amazon EC2, and Amazon Lambda, which show that rateless coding gives as much as 3x speed-up over uncoded schemes.

关键词： Erasure coded computing Rateless Fountain Codes large-scale parallel computing

来源：评论

学校读者我要写书评

暂无评论

A scalable parallel black oil simulator on distributed memory parallel computers

引用

JOURNAL OF COMPUTATIONAL PHYSICS 2015年 301卷 19-34页

作者： Wang, Kun Liu, Hui Chen, Zhangxin Univ Calgary Dept Chem & Petr Engn Calgary AB T2N 1N4 Canada

This paper presents our work on developing a parallel black oil simulator for distributed memory computers based on our in-house parallel platform. The parallel simulator is designed to overcome the performance issues of common simulators that are implemented for personal computers and workstations. The finite difference method is applied to discretize the black oil model. In addition, some advanced techniques are employed to strengthen the robustness and parallel scalability of the simulator, including an inexact Newton method, matrix decoupling methods, and algebraic multigrid methods. A new multi-stage preconditioner is proposed to accelerate the solution of linear systems from the Newton methods. Numerical experiments show that our simulator is scalable and efficient, and is capable of simulating extremely large-scale black oil problems with tens of millions of grid blocks using thousands of MPI processes on parallel computers. (C) 2015 Elsevier Inc. All rights reserved.

关键词： Black oil model parallel simulator Preconditioner large-scale parallel computing

来源：评论

学校读者我要写书评

暂无评论

VERTICAL NATURAL MODES OF THE GRAVEL AGGREGATE IN THE BALLASTED RAILWAY TRACK 4

VERTICAL NATURAL MODES OF THE GRAVEL AGGREGATE IN THE BALLAS...

引用

IV International Conference on Particle-based Methods (PARTICLES 2015)

作者： Aikawa, Akira Railway Tech Res Inst Railway Dynam Div Hikari Cho Kokubunji Tokyo 1858540 Japan

ISBN: (纸本)9788494424472

This research investigates the natural vibration characteristics of the ballast layer by using field measurement, full-scale experiments and large-scale finite element analysis. The results indicate that the natural frequency of the vertical elastic vibration of the ballast layer is numerically detected at around 310 Hz at which the whole ballast aggregate repeats the vertical expansion and shrinkage elastically, and that the rigid-body natural vibration numerically occurs approximately at 120 Hz at which the mass of the track structure vibrates simultaneously up and down as depending on the stiffness of the ballast layer. The stress acting on an angularity part of the ballast gravel is inferred to be about 1100 times greater than the average stress on the bottom surface of the sleepers.

关键词： Ballasted Track Natural Frequency large-scale parallel computing FEM

来源：评论

学校读者我要写书评

暂无评论

OPTIMIZATION OF A parallel CFD CODE AND ITS PERFORMANCE EVALUATION ON TIANHE-1A

引用

computing AND INFORMATICS 2014年第6期33卷 1377-1399页

作者： Che, Yonggang Zhang, Lilun Xu, Chuanfu Wang, Yongxian Liu, Wei Wang, Zhenghua Natl Univ Def Technol Sci & Technol Parallel & Distributed Proc Lab Changsha 410073 Hunan Peoples R China

This paper describes performance tuning experiences with a parallel CFD code to enhance its performance and flexibility on large scale parallel computers. The code solves the incompressible Navier-Stokes equations based on the novel Slightly Compressible Model on three-dimensional structure grids. High level loop transformations and argument based code specialization are utilized to optimize its uniprocessor performance. Static arrays are converted into dynamically allocated arrays to improve the flexibility. The grid generator is coupled with the flow solver so that they can exchange grid data in the memory. A detailed performance evaluation is performed. The results show that our uniprocessor optimizations improve the performance of the flow solver for 1.38x to 3.93x on Tianhe-1A supercomputer. In memory grid data exchange optimization speeds up the application startup time by nearly two magnitudes. The optimized code exhibits an excellent parallel scalability running realistic test cases. On 4 096 CPU cores, it achieves a strong scaling parallel efficiency of 77.39% and a maximum performance of 4.01 Tflops.

关键词： Computational fluid dynamics slightly compressible model large-scale parallel computing uniprocessor optimizations in memory grid exchange scalability efficiency

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：