Summary of "All Science is Computer Science" article

The main point of this article is that all branches of science depend on 
some type of computation.  Supercomputers allow the capability of 
sophisticated computation and simulation that could not be accomplished in 
the real world.  They deal with applications in all areas of science, 
from biology to astrophysics to nuclear chemistry, even fields in the 
social sciences, such as anthropology.  There are simulation software 
packages that can animate complex processes such as the big bang, or model 
complicated molecule structures.  These software applications allow the 
researcher to tweak and manipulate parameters to a level that is 
impossible in the real world.  For many applications, the volume of data 
continues to increase immensely, thus rendering supercomputer calculation 
a necessity.  Some computations can require analyzing trillions of 
bytes of data.  Use of supercomputers has also allowed scientists to 
discover flaws in existing standard models.  For example, a computerized 
genome map suggests that there may not be as many genes as previously 
believed.  This finding may lead to the need to reconsider how genetic 
sequences influence the organization of life.  Many of the sciences 
now rely heavily on computational simulation, which has allowed scientists 
to conduct research which was never before possible.  No matter which 
scientific process is under study, it will inevidably be driven by some 
form of computation.


Summary of computational research project at ORNL

This is a project in the field of computational nuclear astrophysics 
where the goal is to develop a model to simulate core collapse supernovae.  
The idea is to understand the explosion mechanism, nucleosynthesis, and 
the process of producing many of the elements in the universe, which are 
essential for life to exist.  Supernovae are extremely energetic 
explosions, and they release energy in the form of neutrinos.  
Observatories will be able to detect and track neutrino events, which can 
serve as a basis for developing supernova models.  Another tool for 
building a supernova model will be gravitational wave observatories which 
will give insight into supernova convection, believed to influence the 
explosion mechanism.  The TeraScale project will make it possible to 
consider multidimensional simulations, which are necessary for 
understanding the explosion mechanism.  The project will require 
visualization and algorithms for solving large linear systems and large 
eigenvalue problems.  It will depend on collaboration of astrophysicists, 
nuclear physicists, applied mathematicians, and computer scientists.