检索结果-内蒙古大学图书馆

International Workshop on Extreme Scale Programming Models and Middlewar (ESPM2)

作者： Valentin Le Fèvre Tetsuzo Usui Marc Casas Barcelona Supercomputing Center Barcelona Spain Next Generation Technical Computing Unit Fujitsu Limited Kawasaki Japan Barcelona Supercomputing Center Universitat Politècnica de Catalunya (UPC) Barcelona Spain

ISBN: (纸本)9781665463409

Sparse linear algebra routines are fundamental building blocks of a large variety of scientific applications. Direct solvers, which are methods for solving linear systems via the factorization of matrices into products of triangular matrices, are commonly used in many contexts. The Cholesky factorization is the fastest direct method for symmetric and positive definite matrices. This paper presents selective nesting, a method to determine the optimal task granularity for the parallel Cholesky factorization based on the structure of sparse matrices. We propose the Opt-D algorithm, which automatically and dynamically applies selective nesting. Opt-D leverages matrix sparsity to drive complex task-based parallel workloads in the context of direct solvers. We run an extensive evaluation campaign considering a heterogeneous set of 35 sparse matrices and a parallel machine featuring the A64FX processor. Opt-D delivers an average performance speedup of 1.75× with respect to the best state-of-the-art parallel methods to run direct solvers.

关键词： Context Linear systems Symmetric matrices Heuristic algorithms Linear algebra Programming Parallel machines

来源：评论

学校读者我要写书评

暂无评论

Modeling and computation of casting process by particle method 73

Modeling and computation of casting process by particle meth...

引用

73rd World Foundry Congress: Creative Foundry, WFC 2018

作者： Kazama, Masaki Suwa, Tamon Maeda, Yasuhiro Application development div. Next generation technical computing unit FUJITSU Limited Japan Department of Mechanical Engineering Daido University Japan

ISBN: (纸本)9788390430638

In this work, the SPH method was applied to the casting, then it is shown that the qualitative tendency is sufficiently expressed by SPH results. In the congress, we will report the detail of these results and numerical method, because we saw the experimental data and our numerical results are show good agreement. Then we would like to show the utility of this simulation by comparing them with the experimental results. © 73rd World Foundry Congress "Creative Foundry", WFC 2018 - *** Rights Reserved.

关键词： Numerical methods

来源：评论

学校读者我要写书评

暂无评论

Application of explicit and implicit smoothed particle hydrodynamics simulation to casting processes 73

Application of explicit and implicit smoothed particle hydro...

引用

73rd World Foundry Congress: Creative Foundry, WFC 2018

作者： Suwa, Tamon Kazama, Masaki Maeda, Yasuhiro Fujitsu Limited Next Generation Technical Computing Unit 4-1-1 Kamikodanaka Nakahara-ku Kawasaki Japan Daido University Department of Mechanical Engineering 10-3 Takiharu-cho Minami-ku Nagoya Japan

ISBN: (纸本)9788390430638

We simulated the casting processes using SPH method and compared them with the experiments. The results reproduce the experiments well, which suggests advantage of the features of SPH method. It was also found that the implicit method gives better results about oscillation dissipation. © 73rd World Foundry Congress "Creative Foundry", WFC 2018 - *** Rights Reserved.

关键词： Hydrodynamics

来源：评论

学校读者我要写书评

暂无评论

The Tofu Interconnect D

The Tofu Interconnect D

引用

IEEE International Conference on Cluster computing

作者： Yuichiro Ajima Takahiro Kawashima Takayuki Okamoto Naoyuki Shida Kouichi Hirai Toshiyuki Shimizu Shinya Hiramoto Yoshiro Ikeda Takahide Yoshikawa Kenji Uchida Tomohiro Inoue Next Generation Technical Computing Unit Fujitsu Limited Kawasaki Japan AI Platform Business Unit Fujitsu Limited Kawasaki Japan

In this paper, we introduce a new and highly scalable interconnect called Tofu interconnect D that will be used in the post-K machine. This machine will officially be operational around 2021. The letter D represents high “density” node and “dynamic” packet slicing for “dual-rail” transfer. Herein we describe the design and the evaluation results of TofuD. Due to the high-density packaging, the optical link ratio of TofuD has decreased to 25% from the 66% optical link ratio of Tofu2. TofuD applies a new technique called dynamic packet slicing to reduce latency and to improve fault resilience. The evaluation results show that the one-way 8-byte Put latency is 0.49 μs. This is 31% lower than the latency of Tofu2. The injection rate per node is 38.1 GB/s which is approximately 83% of the injection rate of Tofu2. The link efficiency is as high as approximately 93%.

关键词： Bandwidth Optical fiber communication High-speed optical techniques Packaging Network interfaces Resilience Throughput

来源：评论

学校读者我要写书评

暂无评论

Parallel loop cluster quantum Monte Carlo simulation of quantum magnets based on global union-find graph algorithm

arXiv

引用

arXiv 2018年

作者： Todo, Synge Matsuo, Haruhiko Shitara, Hideyuki Department of Physics University of Tokyo Tokyo113-0033 Japan Institute for Solid State Physics University of Tokyo Kashiwa277-8581 Japan Application Research and Development Div. Next Generation Technical Computing Unit Fujitsu Limited Kawasaki211-8588 Japan

A large-scale parallel loop cluster quantum Monte Carlo simulation is presented. On 24,576 nodes of the K computer, one loop cluster Monte Carlo update of the world-line configuration of the S = 1=2 antifer-romagnetic Heisenberg chain with 2.6 × 106spins at inverse temperature 3.1 × 105is executed in about 8.62 seconds, in which global union-find cluster identification on a graph of about 1.1 trillion vertices and edges is performed. By combining the nonlocal global updates and the large-scale parallelization, we have virtually achieved about 1013-fold speed-up from the conventional local update Monte Carlo simulation per-formed on a single core. We have estimated successfully the antiferromagnetic correlation length and the magnitude of the first excitation gap of the S = 4 antiferromagnetic Heisenberg chain for the first time as ξ = 1.040(7) × 104and Δ = 7.99(5) × 10-4, respectively. Copyright © 2018, The Authors. All rights reserved.

关键词： Monte Carlo methods

来源：评论

学校读者我要写书评

暂无评论

SPARC64™ XIfx: Fujitsu's next generation processor for HPC 26

SPARC64™ XIfx: Fujitsu's next generation processor for HPC

引用

26th IEEE Hot Chips Symposium, HCS 2014

作者： Yoshida, Toshio Next Generation Technical Computing Unit Fujitsu Limited Japan

This article consists of a collection of slides from the author's conference presentation on the special features, system design and architectures, processing capabilities, and targeted markets for Fujitsu's S...

ISBN: (纸本)9781467388832

关键词：

来源：评论

学校读者我要写书评

暂无评论

SPARC64™ XIfx: Fujitsu's next generation processor for HPC

SPARC64™ XIfx: Fujitsu's next generation processor for HPC

引用

IEEE Hot Chips Symposium (HCS)

作者： Toshio Yoshida Next Generation Technical Computing Unit Fujitsu Limited

ISBN: (纸本)9781467388849

关键词： Computer architecture next generation networking Microarchitecture Servers Supercomputers Program processors

来源：评论

学校读者我要写书评

暂无评论

The Tofu Interconnect 2

The Tofu Interconnect 2

引用

IEEE Symposium on High Performance Interconnects

作者： Yuichiro Ajima Tomohiro Inoue Shinya Hiramoto Shun Ando Masahiro Maeda Takahide Yoshikawa Koji Hosoe Toshiyuki Shimizu Next Generation Technical Computing Unit Fujitsu Limited Kawasaki Japan

The Tofu Interconnect 2 (Tofu2) is a system interconnect designed for the successor model of the FUJITSU Supercomputer PRIME HPC FX10. Tofu2 uses a 25 Gbps transmission technology that is about two times faster than those of existing HPC interconnects, and uses optical transceivers in an extremely high ratio. Despite the major change of physical transport medium, Tofu2 inherits and enhances the features of Tofu1. This paper describes the specifications including frame format, implementation, and preliminary evaluation results of Tofu2. The effective throughput of Put transfer was evaluated to be 11.46 GB/s that was about 92% link efficiency. The total throughput of simultaneous 4-way transfer was evaluated to be 45.82 GB/s. Tofu2 reduced one-way communication latency by about 0.2 usec. The elimination of the host bus contributed about 60 nsec of reduction, and the rest of the reduction had been derived from the cache injection technique.

关键词： Throughput Optical interconnections Transceivers Data transfer Integrated optics Optical switches

来源：评论

学校读者我要写书评

暂无评论

First-principles calculations of electron states of a silicon nanowire with 100,000 atoms on the K computer

First-principles calculations of electron states of a silico...

引用

2011 International Conference for High Performance computing, Networking, Storage and Analysis, SC11

作者： Hasegawa, Yukihiro Iwata, Jun-Ichi Tsuji, Miwako Takahashi, Daisuke Oshiyama, Atsushi Minami, Kazuo Boku, Taisuke Shoji, Fumiyoshi Uno, Atsuya Kurokawa, Motoyoshi Inoue, Hikaru Miyoshi, Ikuo Yokokawa, Mitsuo Next-Generation Supercomputer R and D Center Development Group RIKEN Japan Center for Computational Sciences University of Tsukuba Japan Department of Applied Physics School of Engineering University of Tokyo Japan Computational Science and Engineering Solution Division Technical Computing Solution Unit Fujitsu Limited Japan PA Project Next Generation Technical Computing Unit Fujitsu Limited Japan

ISBN: (纸本)9781450307710

Real space DFT (RSDFT) is a simulation technique most suitable for massively-parallel architectures to perform first-principles electronic-structure calculations based on density functional theory. We here report unprecedented simulations on the electron states of silicon nanowires with up to 107,292 atoms carried out during the initial performance evaluation phase of the K computer being developed at RIKEN. The RSDFT code has been parallelized and optimized so as to make effective use of the various capabilities of the K computer. Simulation results for the self-consistent electron states of a silicon nanowire with 10,000 atoms were obtained in a run lasting about 24 hours and using 6,144 cores of the K computer. A 3.08 peta-flops sustained performance was measured for one iteration of the SCF calculation in a 107,292-atom Si nanowire calculation using 442,368 cores, which is 43.63% of the peak performance of 7.07 peta-flops. Copyright 2011 ACM.

关键词： Density functional theory

来源：评论

学校读者我要写书评

暂无评论

First-principles calculations of electron states of a silicon nanowire with 100,000 atoms on the K computer

First-principles calculations of electron states of a silico...

引用

Supercomputing Conference

作者： Yukihiro Hasegawa Jun-Ichi Iwata Miwako Tsuji Daisuke Takahashi Atsushi Oshiyama Kazuo Minami Taisuke Boku Fumiyoshi Shoji Atsuya Uno Motoyoshi Kurokawa Hikaru Inoue Ikuo Miyoshi Mitsuo Yokokawa Development Group Next-Generation Supercomputer Research and Development Center RIKEN Japan Center for Computational Sciences University of Tsukuba Japan Computational Science and Engineering Solution Division Technical Computing Solution Unit Fujitsu Limited Japan PA Project Next Generation Technical Computing Unit Fujitsu Limited Japan

关键词： Computers Wave functions Computational efficiency Computational modeling Extraterrestrial measurements Atomic measurements Silicon

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：