[Bioclusters] Question about grid

Karl Podesta bioclusters@bioinformatics.org
Mon, 17 May 2004 19:07:31 +0100


On Thu, May 13, 2004 at 02:41:17PM +0100, Tim Cutts wrote:
> The only reason Seti@home succeeded as a grid computing project is that 
> the compute/IO ratio is *huge*.  Genomics applications just can't do 
> this, although conceivably some things such as molecular dynamics 
> simulations might.

(I hate to do this, but it is an important distinction as different 
concepts and solutions are involved).

SETI@home is a distributed computing project - not a Grid project. 
It is essentially a single program, special-purpose, and many people
choose to run it on their computers. A Grid is a network, or an
infrastructure (rather than a program), so it's a different sort of concept.

Research in the areas obviously overlap, but it's more helpful to 
think of Grid as the network, and of distributed computing projects
(SETI@home, Folding@home) as programs. Grid research is more about 
building infrastructure (computer science) than it is about an optimal 
implementation of your science on a computer (computational science).

Argh, pedanticism and terminology. But this is the general gist. (?)

Kp
--
Karl Podesta
School of Computing
Dublin City University, Ireland