CZ4102 High Performance Computing (Essential, 4MC)

Synopsis
Prerequisite: CZ2102
Introduction to high-performance computing, including comparison of parallel architecture (vector/distributed) and software environments (threads vs. MPI, performance analysis of programs, algorithms in scientific computations (matrix multiplication, linear systems), use of parallel softwarelibraries.

Instructor
Dr. Chen Yu Zong
Department of Computational Science
National University of Singapore
Office: Blk S17 Room 07-24
Tel: 6874-6877. Fax: 6774-6756
E-mail: yzchen@cz3.nus.edu.sg
Web: http://www.cz3.nus.edu.sg/~yzchen (Lots of info about biocomputing)

Schedule

• Lectures:
  • LT22. Mon 14:00-16:00
• Labs:
  • CZ Y3 Lab (S17-07-31)
• Tutorials:
  • S17-07-41. Mon,Tue 10:30-12:00
• Grade:
  • One final = 40%,
  • One mid-term = 30%,
  • Lab = 30%
• Exams:
  • One mid-term exam:
  • One final exam: (12 Nov Monday night)

Module Outline

• Introduction.
• Using MPI: Basic Concepts.
• Using MPI: Sending and Receiving messages.
• Using MPI: How to design a parallel computation by using MPI?
• Using MPI: Nonblocking Send and Receive.
• Using MPI: Collective Communications in MPI.
• Using MPI: Communicators and Groups.
• Using MPI: Optimization.

Lab Schedule

• Lab 1: MPI programing I: Matrix operation.
• Lab 2: MPI programing II: Integral.
• Lab 3: MPI programing III: MPI groups and communicator management.

Tutorials

• Tutorial 1: MPI basics.
• Tutorial 2: MPI routines.
• Tutorial 3: Sample exam questions.
• Tutorial 4: Enrichment topics.

Textbook
Using MPI, 2nd Edition. William Gropp, Ewing Lusk, and Anthony Skjellum, published by MIT Press ISBN 0-262-57132-3. Example programs of this book can be downloaded here.

References:

• Designing and Building Parallel Programs. ISBN 0-201-57594-9; Published by Addison-Wesley. Online material about this book.
• Parallel Programming With MPI. Peter S. Pacheco, published by Morgan Kaufmann.
• On Line Slides: Writing Message-Passing Parallel Programs with MPI.
• A draft of a Tutorial/User's Guide for MPI by Peter Pacheco .