Parallel Computing
TeachersNISHIKI Shinnosuke
Grade, SemesterYear 1 full year [Division of Informatics Science〈Correspondence Course〉(Master's Degree Program)]
CategorySpecial Subjects
Classesメディア授業
Elective, CreditsElective 2credit
 Syllabus Number

Course Description

In recent decades, the development of computer technology has been remarkable, and accordingly, the request for executing large-scale numerical simulations has increased. In order to decrease the calculation time, parallel computing is employed. There are roughly two types of parallel computing: shared memory parallel and distributed memory parallel. You will study the concepts and features of each method, and how to program and execute parallel computing using MPI or OpenMP. In addition, performance of parallel computation of simple fluid analysis will be performed using open source thermal-fluid analysis software, and performance will be evaluated.
You will acquire knowledge and skills on DP1 and DP3 of diploma policies.

Course Objectives

The goal is to understand the concepts and techniques of distributed computing. So that you will become able to create parallelized programs using MPI or OpenMP for simple problems, and execute parallel computation on a computer.

Grading Policy

Your grade will be assessed based on the scores of reports of exercise problems (80%) and final exams (20%). It should be noted that four exercise problems will be given.
LMS posts class materials, receives reports, and provides feedback.

Textbook and Reference

KindTitleAuthorPublisher
Textbook
ReferencesIt is recommending that you collect information using library materials and internet.
Referencesスパコンプログラミング入門: 並列処理とMPIの学習片桐孝洋東京大学出版会、ISBN-13: 978-4130624534
References並列プログラミング入門: サンプルプログラムで学ぶOpenMPとOpenACC片桐孝洋東京大学出版会、ISBN-13:978-4130624565

Requirements(Assignments)

For the preparation, you should study in advance for each class based on reference books, library materials, and materials collected via the Internet. (2 hours)
After the class, you should review and work on exercise problems for report in order to improve your understanding. (1 hour)

Note

Open source compiler and thermal-fluid analysis software will be installed on your computer and calculations will be executed.

Schedule

1Proceeding of classes, history of computers, overview of parallel computing
2C compiler setup and programming basics
3Fortran compiler setup and programming basics
4Overview of shared memory parallel and distributed memory parallel, and parallel computing
5OpenMP for Shared Memory Parallel Processing (1): Overview
6OpenMP for Shared Memory Parallel Processing (2): Programming for parallelization
7OpenMP for Shared Memory Parallel Processing (3): Execute of Parallel computing
8OpenMP for Shared Memory Parallel Processing (4): Performance evaluation
9MPI for Distributed Memory Parallel Processing (1): Overview
10MPI for Distributed Memory Parallel Processing (2): Programming for parallelization
11MPI for Distributed Memory Parallel Processing (3): Execute of Parallel computing
12MPI for Distributed Memory Parallel Processing (4): Performance evaluation
13Fire Dynamics Simulator (FDS) (1): Overview and setup
14Fire Dynamics Simulator (FDS) (2): Creating an input file
15Fire Dynamics Simulator (FDS) (3): Performance evaluation of parallelization using thermal-fluid analysis