2019-A-NK

國家理論科學中心數學組「高效能計算 」短期課程

Introduction to Parallel Programming for Multicore/Manycore Clusters

2019年2月11日星期ㄧ至2月14日星期四 (9:00-17:00),國立臺灣大學 天文數學館 440 室

快速累積的大量數據,以及高速發展的超級電腦,是當今與未來的重要趨勢,也是人工智慧的重要基礎。具備高速計算能力,才能解決更大更複雜的問題,也能大幅提升進行尖端研究與開發產業應用的競爭力。在這個四天的密集課程,我們將簡介 MPI 與 OpenMP 的平行計算環境,說明如何在此平行環境求解稠密矩陣的特徵值問題,並將有限體積法以及大型線性系統疊代法平行化,求解三維 Poisson 方程。課程中的實機操作,將使用當今最先進的超級電腦 Readbush-U supercomputer at the University of Tokyo (附註:帳號須經申請,並依相關規定審核通過後,方得使用)。這個短期課程提供一個非常難得的機會,可以接觸到目前全世界最尖端的高速平行計算環境。歡迎教師、研發人員、碩博士生、大學部同學報名參加。本課程以英語講授。

請在 2019/01/16 前線上註冊 : https://goo.gl/forms/e204zTXS1kx1OiG72

(本課程名額有限,獲得錄取學員將另行通知。)

最新訊息 (News)


直播與錄影 (Online Streaming and Video On Demand)

課程講義 (Course Handouts)

課程簡介 (Overview)

In order to make full use of modern supercomputer systems with multicore/manycore architectures, hybrid parallel programming with message-passing and multithreading is essential. MPI for message-passing and OpenMP for multithreading are the most popular ways for parallel programming on multicore/manycore clusters.

This 4-day tutorial provides essential knowledge and experiences for parallel programming using MPI and OpenMP. Hands-on exercise by the Readbush-U supercomputer at the University of Tokyo (Intel Broadwell-EP) is also given. (https://www.cc.u-tokyo.ac.jp/system/reedbush/index-e.html).

In this course, OpenMP, MPI and OpenMP/MPI Hybrid are applied to 3D Poisson equation solver by finite-volume method (FVM) with preconditioned conjugate gradient iterative method (PCG). Detailed lectures on data structure for parallel FVM, as well as fundamental trainings for MPI and OpenMP are also provided.

The 4th day of the course will be spent for lectures and hands-on on more advanced topics, such as automatic tuning (AT), and large-scale parallel computing.

預備知識 (Prerequisites)

  • Experiences in Unix/Linux
  • Experiences in emacs or vi
  • Experiences of programming (Fortran or C/C++)
  • Fundamental numerical algorithms (Gaussian Elimination, LU Factorization, Jacobi/Gauss-Seidel/SOR Iterative Solvers)
  • Experiences in SSH Public Key Authentication Method

相關軟體 (Software)

授課教授 (Instructors)

Prof. Kengo Nakajima

中島 研吾 教授,東京大學 情報基盤中心 超級計算研究部門

Prof. Takahiro Katagiri

片桐 孝洋 教授,名古屋大學 情報基盤中心 大規模計算支援環境研究部門

課程內容 (Schedule and Contents)

February 11, 2019 (Mon)

09:10-10:00 Introduction

10:10-11:00 FVM code (1/3)

11:10-12:00 FVM code (2/3)

13:10-14:00 FVM code (3/3)

14:10-15:00 Using the Reedbush-U

15:10-16:00 Parallel FVM (OpenMP) (1/3)

16:10-17:00 Parallel FVM (OpenMP) (2/3)


February 12, 2019 (Tue)

09:10-10:00 Parallel FVM (OpenMP) (3/3)

10:10-11:00 Introduction to Parallel Programming

11:10-12:00 MPI Programming (1/7)

13:10-14:00 MPI Programming (2/7)

14:10-15:00 MPI Programming (3/7)

15:10-16:00 MPI Programming (4/7)

16:10-17:00 MPI Programming (5/7)


February 13, 2019 (Wed)

09:10-10:00 MPI Programming (6/7)

10:10-11:00 MPI Programming (7/7)

11:10-12:00 Parallel Data Structure (1/2)

13:10-14:00 Parallel Data Structure (2/2)

14:10-15:00 Parallel FVM (OpenMP/MPI) (1/5)

15:10-16:00 Parallel FVM (OpenMP/MPI) (2/5)

16:10-17:00 Parallel FVM (OpenMP/MPI) (3/5)


February 14, 2019 (Thu)

09:10-10:00 Parallel FVM (OpenMP/MPI) (4/5)

10:10-11:00 Parallel FVM (OpenMP/MPI) (5/5)

11:10-12:00 Advanced Topics (1/5)

13:10-14:00 Advanced Topics (2/5)

14:10-15:00 Advanced Topics (3/5)

15:10-16:00 Advanced Topics (4/5)

16:10-17:00 Advanced Topics (5/5)


主辦單位科技部 國家理論科學研究中心數學組

協辦單位東京大學情報基盤中心科技部人工智慧技術暨全幅健康照護聯合研究中心台灣工業與應用數學學會臺灣大學數學系與應用數學科學研究所

主持人:王偉仲 (台灣大學應數所),林文偉 (交通大學應數系),舒宇宸 (成功大學數學系),黃聰明 (台灣師範大學數學系)

聯絡人:游墨霏小姐 (02-3366-8814, murphyyu@ncts.ntu.edu.tw)