Biomedical Engineering Reference
In-Depth Information
15.1
INTRODUCTION
According to Wikipedia ( http://en.wikipedia.org/wiki/Grid_computing ), “ dis-
tributed” or “grid” computing in general is a special type of parallel computing
that relies on complete computers [with onboard central processing units
(CPUs), storage, power supplies, network interfaces, etc.] connected to a network
(private, public, or the Internet) by a conventional network interface, such as
Ethernet. This is in contrast to the traditional notion of a supercomputer, which
has many processors connected by a local high-speed computer bus.
Grid technology promises to revolutionize many services already offered
by the Internet because it offers rapid computation, large-scale data storage,
and fl exible collaboration by harnessing together the power of a large number
of commodity computers or clusters of other basic machines. The grid was
devised for use in scientifi c fi elds, such as particle physics and bioinformatics,
in which large volumes of data or very rapid processing or both are necessary
[1, 2]. Since 2002, our group has been exploring how to use grids for the life
sciences. We have seen the emergence of the technology in the DataGrid and
Enabling Grids for E-science (EGEE) projects [3], and we are now witnessing
the emergence of production grids, called e-infrastructures, at national and
European scales.* It is still not easy to use grids for someone who is not famil-
iar with Linux; however, the quality of service is considerably better than eight
years ago. One can compare the present situation to a glass which is half full:
Looking at the fi lled half, one can appreciate the capacity to access tens of
thousands of cores on demand and now have robust data management tools.
Looking at the empty half, one can still complain about the lack of advanced
data management service and the limited user friendliness. But everybody will
agree that (at the time of writing in August 2010) the existing grid services are
better than they have ever been. In other words, grids offer more opportunities
than ever to their users to do science differently.
The goal of this chapter is to explain how grids can be used to address the
grand scientifi c challenges of the twenty-fi rst century in a more innovative and
collaborative way. We will provide examples from our experience over the last
eight years to show how grids can be used today by researchers. We will also
discuss how new technologies are emerging and are going to enrich the ser-
vices offered to the grid users.
15.2
GRIDS FOR E - SCIENCE
Two trends are irreversible in science: First, more and more scientifi c data are
produced and must be analyzed. Examples are easy to fi nd in almost every
fi eld of science. In high-energy physics, chasing for new particles at the large
hadron collider requires selecting a few events per year out of billions taking
*
See the website of the e-infrastructure Refl ection Group and references therein at http://www.e-
irg.eu/ .
Search WWH ::
Custom Search