TAILIEUCHUNG - Báo cáo " Application of a Godunov type numerical scheme and a domain decomposition technique to the parallel computation of tidal propagation "

A parallel computing model for the numerical solution of the general 20 shallow water equations in conservative form has been developed, tested and implemented in the MPI parallel environment set up on a parallel computer with four GHz CPUs in Institute of Mechanics, VAST. The model is based on a Godunov-type numerical scheme, which is devised for 20 unstructured computational meshes, and on a domain decomposition technique. | VNU Journal of Science Earth Sciences 25 2009 104-115 Application of a Godunov type numerical scheme and a domain decomposition technique to the parallel computation of tidal propagation Nguyen Tat Thang Nguyen The Hung Institute of Mechanics Vietnam Academy of Science and Technology VAST Received 15 June 2009 received in revised form 23 June 2009 Abstract. A parallel computing model for the numerical solution of the general 2D shallow water equations in conservative form has been developed tested and implemented in the MPI parallel environment set up on a parallel computer with four GHz CPUs in Institute of Mechanics VAST. The model is based on a Godunov-type numerical scheme which is devised for 2D unstructured computational meshes and on a domain decomposition technique. That newly developed parallel computing model has been applied to a tidal wave propagation problem in the Gulf of Tonkin in the South China Sea. The computed results show good agreement with that of the TIDE2D model. It is of high importance that computation time is significantly reduced. Keywords Parallel Computing Godunov Type Numerical Scheme 2D Shallow Water Model Domain Decomposition Technique. 1. Introduction Numerical . methods have long been developed and applied to many scientific problems. The finite volume method FVM which is conservative has recently been applied to fluid dynamics simulations and good results have been obtained 1 . The application of the FVM method in combination with a Godunov type numerical scheme and the approximate Riemann solver of Roe is a prospective approach for the numerical solution of the shallow water equations 2 . 3 . This method is Corresponding author Tel. 44 01224 273519. Email especially appropriate in problems where hydraulic jumps shockwaves or wave propagation may present 4 . A numerical model based on this approach for the solution of the general 2D shallow water equations in conservative form has been developed .

• Báo cáo khoa học
• Parallel Computing
• earth sciences
• Research
• scientific research
• scientific reports
• Lecture Parallel computing & Distributed systems
• Parallel computing & Distributed systems
• Bài giảng Tính toán song song và phân tán
• Parallel computer architectures
• Flynn’s taxonomy
• Conventional sequential machines
• Parallel matrix processing
• OpenMP Core Syntax
• Nested loops
• Introduction to Parallel Computing
• Algorithm design and parallel programming
• Design and analysis of parallel algorithms
• Parallel programming platforms
• Basic communication operations
• Analytical modeling of parallel programs
• Parallel processing
• Financial disaster modeling
• High perfomance computing
• Programming shared address space platforms
• Dense matrix algorithms
• Fast fourier transform
• Cache organization
• Fully set associative
• n way set associative
• Advanced topics
• Amdahl’s law
• Gustafson’s law
• The dRuby Book
• Distributed
• Parallel Computing with Ruby
• Distributed Ruby
• Tính toán song song
• Bài giảng Tính toán song song
• Kiến trúc song song đồng bộ
• Tổ chức các bộ vi xử lý
• Các mảng bộ vi xử lý
• Hệ thống xử lý song song
• Hệ thống tính toán song song
• Hệ thống nhiều máy tính
• Lập trình song song
• Phương pháp trao đổi thông điệp
• Máy ảo song song
• Thuật toán song song
• Mô hình PRAM
• Thuật toán song song nhân hai ma trận
• Xử lý song song
• Xu hướng phát triển của CPU
• Mô hình lập trình song song
• Thuật ngữ tính toán song song
• Bộ nhớ máy tính song song
• Kiến trúc bộ nhớ máy tính song song
• MPJ Express
• Sequential rule
• Association rule
• High performance
• Quantum impurity system
• Time dependent spectral function
• Degrees of freedom
• The corresponding time evolution
• BMC Bioinformatics
• Population genetics
• Graphical user interface
• Linkage disequilibrium
• Information sciences
• Hardware architecture for membrane computing
• Parallel architectures
• Hardware design
• the parallel working mechanisms
• Market Oriented Grid
• Utility Computing
• Wiley Series on Parallel
• Distributed Computing
• market oriented
• High performance computing
• DNA methylation
• Parallel pipeline
• Methylation analysis
• Phân tán dữ liệu
• Sự phát triển máy tính nhanh nhất
• Xu hướng máy tính
• Tính toán hiệu nâng cao
• Mô hình của MPI
• NGS data analysis
• RNA Seq
• Cloud computing
• Paired end read analysis
• Single end read analysis
• Pipeline architecture
• RNA folding
• Parallel biological computing
• Loop tiling
• Transitive closure
• Loop skewing