SIAM Short Course on the Portable, Extensible Toolkit for Scientific computation (PETSc)

Sunday, March 21, 1999
9:00 AM-5:30 PM
Room: Executive Salon 1, third floor
Adam's Mark San Antonio-Riverwalk Hotel
San Antonio, Texas

Held in conjunction with the
Ninth SIAM Conference on Parallel Processing for Scientific Computing

Contents

Questions? E-mail meetings@siam.org

  SIAM Conferences

 

Organizers and Instructors

Satish Balay, William Gropp, Lois Curfman McInnes, and Barry Smith, Mathematics and Computer Science Division, Argonne National Laboratory

Rationale

The development of application codes to simulate physical phenomena is complex, time-consuming, and requires specialized expertise. Developing parallel codes from scratch is often beyond the resources of any individual or small team. The PETSc toolkit provides a rich set of data structures and algorithms to ease the development of scalable PDE-based simulation codes. PETSc provides, not only a variety of parallel linear and nonlinear algebraic solvers, but, also much of the software infrastructure to manage the grid data structures without requiring one to program explicitly in MPI.

Description

This course introduces and motivates the organization and use of PETSc for the parallel solution of PDEs.

Level of Material

  • Introductory 30%

  • Intermediate 40%
  • Advanced 30%

Course Objectives

Attendees will be able to write parallel codes using PETSc.

Who Should Attend

Scientists, engineers, and mathematicians who are interested in the large-scale solution of PDEs.

Recommended Background

The attendee should have experience in programming in Fortran, C, or C++ and knowledge and experience in the numerical solution of PDEs. Experience with MPI or parallel computing is desirable but not necessary.

Instructors

Satish Balay received his Masters in Computer Science from Old Dominion University in 1995.

William Gropp received his Ph.D. in Computer Science from Stanford University in 1982.

Lois Curfman McInnes received her Ph.D. in Applied Mathematics from the University of Virginia in 1993.

Barry Smith received his Ph.D. in Mathematics from New York University in 1990.

All four have extensive experience in parallel programming for numerical computations. PETSc was begun in 1991 by Gropp and Smith, version 2.0 has been continuously supported since June 1995 and has been used in a variety of parallel application codes.

Additional Information

The PETSc software package is freely available from http://www.mcs.anl.gov/petsc.

Program

Morning

8:00 Registration

9:00-10:00 Introduction

  • What problems is PETSc intended to solve?

  • Sample results from PETSc application codes

10:00-10:30 Coffee

10:30-12:30 Field and grid data management

  • Structured grid codes

  • Unstructured grid codes

Afternoon

12:30-2:00 Lunch

2:00-3:30 Solvers

  • Assembling Jacobian matrices

  • Nonlinear solvers
  • Linear solvers
  • ODE solvers

3:30-4:00 Coffee

4:00-5:30 Putting it all together

  • Profiling/performance tuning

  • Debugging

5:30 Short course adjourns

Location

The short course on PETSc will take place in Executive Salon 1 on the third floor. Coffee breaks will be in Lobby, third floor and lunch will be in the Rose Garden Room.

How to Register

Seats are limited. We urge short course participants to register in advance. To register, please complete the Preregistration Form online and send it with your payment to reach SIAM office on or before March 3, 1999.

Registration fee for the short course includes: course notes, final program and abstracts, coffee breaks, and lunch on Sunday, March 21.

PP99 Home

Program

Program Updates

Speaker Index

Hotel

Transportation

Registration

Maintained by Maryann M. Donaghy, donaghy@siam.org
Created: TJF, 10/20/98 MMD, Updated: 1/29/99