Once each semester the CAC offers short classes on several topics. The classes are generally about 2 hours long and are held on North Campus. While the classes are introductory in nature, some familiarity with Unix or Linux systems is very helpful.

Current Sessions Offered

Introduction to MPI
High Performance Mathematics

Past Sessions

MATLAB for Research Computing
Handouts and Examples

High Performance Math - BLAS, FFT, and LAPACK

Topics will include:

• Why writing code is bad
• Extracting performance using libraries
• The Basic Linear Algebra Subprograms BLAS
• Linear Algebra Package LAPACK
• Free Parallelism with threaded BLAS
• FFT's using FFTW

Introduction to MPI Programming

Topics will include the ideas behind parallel programming, an introduction to the Message Passing Interface functions, and some common mistakes.

MATLAB for Research Computing

Topics will include mcc the MATLAB compiler, parallel MATLB, running MATLAB in batch, MATLAB performance profiling, and MATLAB licensing issues.

Documents and Sample Code

Below is an auto generated list of PDF's of all the finished training sessions. Introduction training is offered three times a year for all users. Research groups of 5 or more, just getting started, may request training outside the normal schedule. Please contact cac-support@umich.edu to schedule.

New training sessions are posted on the front page

High Performance Math
./blas-code.tar.gz
An Introduction to The Center for Advanced Computing
./cac-intro-code.tar.gz
An Introduction to MPI
./cac-mpi-code.tar.gz
MATLAB for Research Computing
./matlabcode.tar.gz