-----
Title:
-----

Large Scale and Distributed Computing: 
Application Development Towards Petascale Platforms
In honor of Prof. J. Tinsley Oden's 70th birthday 

-----------
Organizers:
-----------

S. J. Kim 
sjkim@dosa.snu.ac.kr

A. K. Patra
Department of Mechanical and Aerospace Engineering
University at Buffalo, Buffalo, NY 14260
Tel 7166452593x2240

-------------------------
Minisymposium description:
------------------------- 

High fidelity computational simulations of physical systems require the 
development of new methods to use the largest possible memory and processing 
power. Distributed memory computing allows applications access to large 
memory and processing power while overcoming the processor memory bandwidth 
bottleneck but comes with the burden of distributing and sychronizing computations.  
Such machines also have deep memory hierarchies and most recently complex 
multi-core architectures that must be optimally used to run applications 
efficiently. In conjuction with mesh and model adaptivity large scale computing 
will extend our ability to address grand challenge computations by conducting 
simulations in the most efficient manner on the largest possible machines. 
Domain Decomposition equation solvers, Mesh Partitioning for problem 
decompositions, Managing Cache Efficiencies, programming for Multi-Core 
architectures and many other research challenges must be overcome to accomplish
the goal of accurate simulations for many grand challenge applications. Some 
of the above issues have challenged researchers over the last 20 years and 
others are new but much remains to be done to develop successful petascale 
applications. We solicit presentations on all research issues related to 
developing such simulations.