[Bioclusters] Anyone interested in clustering transmeta cpus?

Ivo Grosse grosse@cshl.org
Sat, 25 Aug 2001 13:39:32 -0400

Hi Chris,

thanks for your excellent comments.  

chris dagdigian <dag@sonsorol.org> wrote on Sat, 25 Aug 2001:

> Blackstone and others have tacked this problem by breaking up the databases 
> into pieces that are small enough to be dynamically shipped peer-to-peer 
> It does work though- we did some blazing fast searching on nodes with 256mb 
> RAM using this approach.

Assume a database that fits into 2 GB RAM.  How much faster is your 
BLAST analysis on 8 blades (with 8 x 256 MB RAM and 8 CPUs) as compared 
to a single-Athlon board with 2 GB DDR?

> There is no need for high speed interconnect for bioinformatics and 
> sequence analysis. As you said most of those apps are embarrassingly 
> parallel and most in fact are rate limited by things like RAM and disk I/O.

That's exactly our experience.

> Once you start having researchers who want to do computational chemistry, 
> molecular modeling, QSAR and virtual screening then 

Okay okay okay ... these are totally different topics ... and you are 
absolutely right in what you write.  I did not refer to those topics.