High Performance Applications

Big Red Cluster Support

Getting Started on Big Red

Working on Big Red Primer
To get started on the Big Red cluster.

Migration between LoadLeveller & PBS
Compare LoadLeveller and PBS commands, options, and such. (Mirror image PBS to LL also available).

Compiling and Linking Applications

Compile Switches for IBM XL Compilers Learn about compiler and linker flags you can use to optimize your code, etc.

Parallel MPI applications on Big Red

Using Paralleljob script to submit MPI jobs Have a simple parallel program and don't want to mess with mpirun, etc.? Then use paralleljob script!

Compiling and Submitting Parallel (MPI) Jobs Learn how you can compile your own parallel MPI jobs and run them on Big Red.

TeraGrid Support

Getting Started on TeraGrid
Learn about how you can get an account and run jobs on the TeraGrid, and so forth.

TeraGrid Knowledge Base
Check out the TeraGrid KB for answers to commonly asked questions.

TeraGrid UserInfo Documentation
The main documentation for TeraGrid users.

Cluster Support

Parallel (MPI) Programming
To learn how you can write parallel programs, compile and run them, debug and profile them.

About us:

Indiana University's high performance computing and communication environment enables researchers across a wide range of academic fields to explore new paradigms in science, the arts, and education. Computing resources are available to the IU community without charge, "just like the library." University Information Technology Services provides hardware, software, networks, and support.

The Libra Cluster provides an extremely robust AIX computing environment supporting a broad range of applications. The Big Red and Quarry Linux clusters make up IU's primary parallel computing environment. They provide the muscle and the tools to manage the massive amounts of data associated with real-time access to scientific instrumentation across the Internet. GRAPEs provide additional computational power. The Research Database Complex serves researchers with large data analysis projects, and the Massive Data Storage System provides over 1.5 petabytes of nearline storage capacity.

Linked by the I-Light optical fiber network directly to the Internet2 and the Global Research Network Operations Center in Indianapolis, these systems are part of a national and international infrastructure forming a bridge for collaboration between Indiana University researchers and their colleagues around the world. Indiana University has been chosen by the NSF to be one of nine member institutions of the TeraGrid.

A researcher at any IU campus wishing to get started in high performance computing should get an account on the Libra or AVIDD cluster. There are good tutorials and other resources available on the internet, and many departments make scientific computing an important part of their programs. Researchers with significant computational projects should get an account on the TeraGrid.

The HPA group provides programming support, for example answering questions about how to gain the advantages such systems can offer. We will help you to migrate, optimize, and parallelize your code. Our group consists of highly qualified professional computational scientists, who seek substantial collaborative relationships with IU research programs. Extended consulting is available. Send questions to hpa-admin at iu.edu.