教員名

Taisuke Boku, Daisuke Takahashi, Osamu Tatebe, Hiroto Tadano, Claus Aranha, Jinpil Lee (RIKEN R-CCS)

所属

Faculty of Engineering, Information and Systems

教務担当

Systems and Information Engineering

主実施組織

Center for Computational Sciences

TA

TF

オフィスアワー

研究室

電話番号

E-mail

High performance computing is the basic technology to support today's large scale scientific simulations. It covers widely spread issues on hardware and software for high-end computing such as high speed computation, high speed networking, large scale memory and disk storage, high speed numerical algorithm, programming scheme and system software to support them. Current advanced supercomputer systems are based on large scale parallel processing systems and it is required even for application users to understand a certain level of these informations for effective utilization of them. In this class, we focus on the basic technology of high-end computing systems, programming, algorithm and performance tuning for application users who aim to use these systems for their practical simulation and computing.

Feb. 3 (Mon)
1　 Fundamentals of HPC and Parallel Processing　 09:00 - 10:30
Taisuke Boku
Amdahl's law, Parallelization methods (EP, Data parallelism, Pipeline parallelism), Communication, Synchronization, Parallelization efficiency, Load balance.

2　 Parallel Processing Systems　 10:45 - 12:15
Taisuke Boku
Parallel processing systems (SMP, NUMA, Cluster, Grid, etc.), Memory hierarchy, Memory bandwidth, Network, Communication bandwidth, Delay.

3　 Parallel Programming 1: OpenMP 13:30 - 15:00
Jinpil Lee (RIKEN R-CCS)
Parallel programming model, parallel programming language OpenMP.

4　 Parallel Programming 2: MPI 15:15 - 16:45
Claus Aranha
Parallel programming language MPI2．

Feb. 4 (Tue)
5　 Optimization 1: Computation Optimization　09:00 - 10:30
Daisuke Takahashi
Program optimization techniques (Register blocking, Cache blocking, Memory allocation, etc.) and performance evaluation on a compute node of parallel processing systems.

6　 Optimization 2: Communication Optimization　 10:45 - 12:15
Osamu Tatebe
Optimization techniques and performance evaluation of parallel programming on parallel processing systems.


7　 Parallel Numerical Algorithm 1　 13:30 - 15:00
Daisuke Takahashi
Fast Fourier Transformation (FFT) and its parallelization methods.

8　 Parallel Numerical Algorithm 2　 15:15 - 16:45
Hiroto Tadano
Krylov subspace iterative methods and their parallelization methods.

The content is the same as 01ZZ606.
For details :
https://www2.ccs.tsukuba.ac.jp/workshop/HPCseminar/2019/lecture_e.html

The class will be taught in 365体育投注.
TWINS registration is available from January 22 (Wed) through February 2 (Sun).