Introduction to Parallel Programming for Multicore/Manycore Clusters 
February 2124, 2017
National Taiwan University, Taipei, Taiwan
Instructor: Takahiro Katagiri (Information Technology Center, Nagoya University,
Japan), and
Kengo Nakajima (Information Technology Center, The University of Tokyo, Tokyo, Japan)
Fundamental (2/2123) (Instructor: Takahiro Katagiri ) 
Trainings of fundamental MPI and OpenMP are
provided with parallelization of dense matrixvector multiplications and power
method for eigenvalue problems.
Fundamental 1/3: February 21th, 2017 (Tuesday)
PDF / Source Codes 
Time 
Contents 
Overview_of_OpenMP.pdf

15:1016:00 
Overview of OpenMP 

16:1017:00 
Functions of OpenMP

Fundamental 2/3: February 22th, 2017 (Wednesday)
PDF / Source Codes 
Time 
Contents 
Training_of_OpenMP.pdf
Sample code for Own PC:
MatMat.tar
Answer code for Own PC:
MatMat_ans.tar

9:1010:00 
Training of OpenMP with own PC

Overview_MPI.pdf

10:1011:00 
Overview of MPI

How_to_use_FX10.pdf 
11:1012:00 
How to use the FX10
(You need a password to access this file.)


13:1014:00 
Training of FX10 


Homework 1
(OpenMP parallelization with the FX10 or
own PC) 
Nonblocking.pdf

14:1015:00 
Functions of MPI Nonblocking and
Persistent Communication 
MatVec.pdf

15:1016:00 
Parallelization of dense MatrixVector
Multiplications (1/2) 

16:1017:00 
Parallelization of dense MatrixVector
Multiplications (2/2) 


Homework 2
(MPI Parallelization of MatrixVector
Multiplication with the FX10)

Fundamental 3/3: February 23rd, 2017 (Thursday)
PDF / Source Codes 
Time 
Contents 
Power_Method.pdf

09:1010:00 
Parallelization of dense Power Method
for eigenvalue problem (1/2) 

10:1011:00 
Parallelization of dense Power Method
for eigenvalue problem (2/2) 


Homework 3
(MPI Parallelization of Power Method with the
FX10)

MatMat_1.pdf

11:1012:00 
Parallelization of Fully Distributed
dense MatrixMatrix Multiplication (1/2) 


Homework 4
(MPI Parallelization of MatrixMatrix Multiplication with the FX10) 
MatMat_2.pdf

13:1014:00 
Parallelization of Fully Distributed dense MatrixMatrix Multiplication (2/2) 


Homework 5
(MPI Parallelization of Fully Distributed MatrixMatrix
Multiplication with the FX10) 
Update: 28th/Jan./2017
