|
Computational Science is the wave of the future. In recent years the balance
between computer simulation and experimentation in Physics, Chemistry, and
the Life Sciences has shifted dramatically with our ability to perform large
scale experiments on larger and larger computer systems. We are at a
threshold in time when the power of computing applied to the Life Sciences
will revolutionize the design of drugs, the understanding and invention of
medical treatment, the analysis of gene expression data and biological
pathway construction. These problems are cellular in nature.
In fact, all of life is cellular. Cellular automata, a subclass of the more
general Finite Element Modelling problems, are natural candidates for large
scale parallel computing systems, such as a computing Grid.
OptimalGrid is a grid computing infrastructure that can address the
special and specific needs of the cellular automata type problems that are
found in the Life Sciences domain.
Furthermore, OptimalGrid targets the fragile and dynamic computing landscape
of the Internet, where computers live and die, go fast and slow, and even
get confused. OptimalGrid manages all aspects of the problem,
such as the dynamic allocation and reallocation of the compute resources, or
the decomposition of a large flow problem into finite elements.
OptimalGrid keeps the large parallel computation running at near optimal
speeds, despite the problems presented by using compute nodes on the
Internet, whose status changes over time.
An Eden Model simulation (game of life)
|
|
