Introduction to Parallel Programming for Multicore/Manycore Clusters (FVM Application)

February 21-24, 2017
Short Course for Scientific Computing (SC4SC 2017)(FVM Application)
National Taiwan University

Instructor: Kengo Nakajima, Ph.D., Information Technology Center, The University of Tokyo

Preparation

• Windows
• Please install Cygwin and ParaView (for visualization) to your PC.
• Moreover, please install "gcc" (C) or "gfortran" (Fortran) in "Devel/Cygwin" and "OpenSSH" in "Net/Cygwin"
• Mac/Unix/Linux
• Please install ParaView (for visualization) to your PC.

General Introduction

• Introduction: Invitation to Supercomputing [click]
• Overview of the Supercomputer Systems in ITC/University of Tokyo [click]
• Technical Terms [click]

Report

• Description [click]

Additional Material(s)

• Tuning/Scalar/Vector Processors [click]

Materials (C)

• Part II-1: FVM Code [click]
• [allocate/deallocate vector/matrix] [click]
• [Programs for PC] [click]
• Part II-2: Parallel FVM using OpenMP [click]
• Part II-3: Parallel FVM using MPI [click]
• [Point-to-Point Communication] [click]
• [Exercise] [click]
• Part II-4: OpenMP/MPI Hybrid [click]
• Part II-5: Communication-Computation Overlapping + Futher Issues [click]

Materials (Fortran)

• Part II-1: FVM Code [click]
• [Programs for PC] [click]
• Part II-2: Parallel FVM using OpenMP [click]
• Part II-3: Parallel FVM using MPI [click]
• [Point-to-Point Communication] [click]
• [Exercise] [click]
• Part II-4: OpenMP/MPI Hybrid [click]
• Part II-5: Communication-Computation Overlapping + Futher Issues [click]