AMATH 583: High-Performance Scientific Computing

SLN 10233, MWF 4:00-4:50, Loew Hall 216 and 206

Instructor:

Professor Randall J. LeVeque
Guggenheim 415C
tel: 685-3037
fax: 685-1440
rjl AT uw DOT edu

TA: Grady Lemoine, gl AT uw DOT edu

Office hours: See Catalyst class webpage

This class meets together with AMath 483.

Course Description

Introduction to hardware, software, and programming for large-scale scientific computing. Overview of multicore, cluster, and supercomputer architectures; procedure and object oriented languages; parallel computing paradigms and languages; graphics and visualization of large data sets; validation and verification; and scientific software development. Prerequisite: either CSE 142 or AMATH 301.

Class notes:

For more details, including an overview of the course and tentative schedule, see the class notes.
These include a draft of Homework 1

Catalyst web page:

Students registered in the class should have access to the Catalyst class webpage. On that page you can find the following:

The Virtual Machine containing required software for downloading
Webstreaming of lectures
Additional sites required for homework problems
Discussion board to get help

Grading:

There will be no exams or large-scale projects in this class. There will be numerous homework assignments aimed at gaining hands on experience with a variety of computational tools and computer languages.

Languages and software:

Homework assignments will require using a variety of software tools and languages, including: Unix, Python, Fortran, OpenMP, MPI, Mercurial, debuggers, and visualization tools. Students are expected to have programming experience at the level of CSE 142, Applied Math 301, or a similar introductory programming course. It is not assumed that students will already be familiar with the languages mentioned above.

The emphasis will be on using open source software that is freely available for a variety of computing platforms. Instructions and some assistance will be available for students who wish to download and install software on their own computers.

As an alternative, a Virtual Machine based on Virtual Box will be available for download that contains all the required software and may be easier to install and use than installing each package separately.