Computational Science at Argonne National Laboratory

Computational science at Argonne has three interrelated aspects: developing innovative computer science tools, software, and algorithms; providing large-scale parallel computing resources; and supporting challenging applications in the physical and biological sciences. Leading these efforts is the Mathematics and Computer Science Division. We welcome both beginning and experienced researchers interested in taking advantage of our expertise in numerical algorithms, tools for parallelism, and other enabling technologies.

What's New?
Argonne is participating in the TeraGrid project to build the world's most powerful computing infrastructure for scientific research.

The Enabling Technology
Research in the Laboratory for Advanced Numerical Software covers the spectrum from numerical algorithm design, analysis, and implementation to the development of parallel programming tools designed to dramatically increase scientific productivity.

PETSc - data structures and routines for solving PDEs
TAO - component-based optimization software
ADIFOR, ADIC - toolkits for generating derivative code
Opt-MS - optimization-based mesh improvement
MPICH - portable, high-performance version of MPI (Message Passing Interface)
Jumpshot - performance visualization
NEOS - optimization solvers on the Web
metaNEOS - metacomputing environments for large-scale optimization
The Resources
Access to world-class computer resources is critical both for developing new computer science technology and for enabling a new generation of computational science applications.
The Advanced Computing Research Facility - parallel computing systems, I/O subsystems, high-capacity external network links, and wireless networking
The Futures Lab - advanced tiled display and collaborative environments
The Distributed Systems Laboratory - research and development of middleware for computational grids
The Science
The ultimate goal of our research is, of course, to enable the solution of problems important to DOE. The Argonne computer resources and MCS-developed software are used in diverse computational science applications. Several other Argonne divisions also have their own computational resources on which they conduct simulations. The following is only a sample of the ongoing computational science research at the Laboratory. For more information, see the individual software projects listed above, and check the Argonne home page for links to the different divisions.

Computational fluid dynamics - fluid mechanics, sediment transfer, biofluids
Computational biology - genome sequence analysis
Regional climate modeling - Midwest climate simulation
Computational astrophysics - visualization, scalable I/O technology
Computational chemistry - chemical kinetics
Nuclear physics - ground and low-lying excited states of light nuclei
Materials science - interfaces of polycrystalline microstructures
Advanced Photon Source -- materials science; biological science; physics; chemistry; and environmental, geophysical, and planetary science.

For information, please contact gropp@mcs.anl .govMCS Logo