[CAnet - news] Amazon EC2 For Scientific Processing

For more information on this item please visit my blog at 
http://billstarnaud.blogspot.com/
-------------------------------------------

[Amazon EC2 is generating a lot of buzz in scientific computing circles as
it makes distributed computing platform projects a lot easier to implement,
as opposed to federating disparate resources from different management
domains. Some also claim that EC2 is considerably cheaper for a researcher
than what it would cost to provide power and cooling to most HPC computing
clusters. Thanks to Richard Ackerman for this pointer-- BSA]

http://aws.typepad.com/aws/2007/06/amazon_ec2_for_.html

Amazon EC2 For Scientific Processing

Bioinformatics_for_dummies Mike Cariaso was kind enough to set up the Meetup
in Bethesda for my upcoming trip to Washington, DC. Mike has done some
pretty cool work with with Amazon EC2, setting up the mpiBLAST tool to run
on EC2.

MPI, short for Message Passing Interface, is a standard for coordinating
processing on supercomputer grids. MPIPCH2 is a popular implementation of
MPI.

BLAST is the primary bioinformatics tool used to query genome sequences
against an established database, or to match one sequence against another.
The primary BLAST tool is run as an online service by the National Institute
of Health.

Running BLAST over MPI lets BLAST run on a processing grid; this variant is
called mpiBLAST.

Mike's work builds on that of Peter Skomorch, who did the work needed to get
MPIPCH2 running on Amazon EC2. Peter documented his work in a very
informative set of blog posts:

    * On-Demand MPI Cluster with Python and EC2 (part 1 of 3)
    * MPI Cluster with Python and Amazon EC2 (part 2 of 3)
    * Amazon EC2 Considered Harmful

That last post doesn't actually reference EC2, but it is entertaining
nonetheless. Part 2 ends with a parallel fractal calculation running on 5
EC2 instances!

By the way, I'm very interested in hearing about more academic and
scientific uses of EC2. Please feel free to post a comment.

-- Jeff;

See also past postings:

http://lists.canarie.ca/pipermail/news/2006/000365.html


-------------------------------------
To SUBSCRIBE:
send a blank e-mail message to
[email protected]

To UNSUBSCRIBE:
send a blank email message to
[email protected]
-------------------------------------

These news items and comments are mine alone and do not necessarily reflect
those  of the CANARIE board or management.
-----------
[email protected]
[email protected]
www.canarie.ca/~bstarn
skype: pocketpro
SkypeIn: +1 614 441-9603


_______________________________________________
news mailing list
[email protected]
http://lists.canarie.ca/mailman/listinfo/news