Hello,

Moderator adds: I am approving the below message as some of you might want to
share your thoughts or point to your efforts on Grid applications. Meanwhile, I
would like to give pointers to two of our documents on Grid applicatoions: 1
Docking on Grid - http://buyya.com/papers/vlab-drug-design.pdf and 2.
Neuroscience on the Grid - http://www.cs.mu.oz.au/~raj/papers/neurogrid-ccpe.pdf
. With this short node, I wish all of your a happy and prosperous new year.
Cheers. Raj

---------------------
Dear all,

I am a newbie to grid computing, and I am seeking some
research topics in this aspect. So I'd like to ask a very
basic question about grid computing, to help my research
direction. Is grid computing suitable for large-scale
scientific computing?

So far I have seen, the existing products and applications
related to grids (esp. with peer-to-peer computing) are mostly
related to file sharing and web search engines. They seem to
share a common feature that the results returned in the search
phase need not be deterministic and 100% accurate. This implies
different memory copies among different nodes need not be always
consistent, hence reducing communication for data propagation
through the WAN, resulting in good performance.

However, scientific computing applications require accurate
and precise results, with little or no margin of error. The
consistency needs to be always obeyed. Does this nature hinder
the use of scientific applications in grids?

Thanks a lot in advance.

Regards,
Benny (bluishlv).