Information Technology Reference
In-Depth Information
Example
The company Sun was among the first movers on the Utility Computing market.
Sun launched the first version of the Sun Grid Compute Utility offering in March
2006 (Schwartz 2006). Users were able to purchase computing capability on a
pay-per-use basis for $1/CPU-hr. Since then Sun has constantly evolved its Utility
computing offerings. For example, besides utility computing Sun currently also
offers support for developers, who are implementing applications that are supposed
to run on the Utility Grid (see also section 4.4.2.1 in chapter 4).
The Sun Grid Compute Utility has been used by organizations from communi-
cations, technology and life sciences industries, e.g. chipmaker Advanced Micro
Devices or genomics R&D group Applied Biosystems (Sun 2009e). Using the Sun
Grid Compute Utility, Applied Biosystems (Life Sciences) was able to perform the
compute-intensive data research to develop millions of new genomic assays in a
matter of days rather than months. In addition, because the company only had to
purchase the number of hours required, at a rate of $1 per CPU hour, it avoided an
investment in infrastructure that would have cost the company hundreds of thou-
sands of dollars.
Another
example for a successful use of Utility Computing that involves not
only computing services, but a combination of computing and application services
is the cooperation of the film studio DreamWorks and HP. HP Labs provided a utility
rendering service using a 1'000-processor data centre that was used by DreamWorks
as a scalable, off-site rendering capacity for the production of high-quality film
animation. The connection between HP Lab's data centre and DreamWorks' studio
20 miles away was established via a secure fibre optic link (HP Labs 2004). The
data centre of HP Labs became a remote extension of DreamWorks' IT infrastruc-
ture. It provided additional compute capacity required for peak periods in the movie
production process.
Organizations can benefit from utility computing offerings like the Sun Grid
Compute Utility or HP Labs' utility rendering service in a twofold way. First, they
can handle compute tasks that exceed the capacity of their own infrastructure, e.g. in
peak times or for specific projects, or drastically reduce processing time of compute-
intensive tasks, e.g. for complex simulations or analyses. Second, they do not need
to make additional infrastructure investments. However, it has to be noted that these
advantages come together with higher security, privacy and operational risks.
3.5.2.4 Partner/Community Grids
The idea of Partner or Community Grids originated from eScience. Many research
endeavours, in particular in natural sciences (see for instance the example of CERN
in chapter 1), require joint research efforts from scientists and sharing of infra-
structures of research institutions from all over the world. The cooperation usually
results in a Virtual Organization (VO) within which resource sharing takes place.
Today, the need for cooperation is increasing in the business world too. Due to
globalization, companies are more and more involved in global supply chains and
the success of a company increasingly depends also on the efficient collaboration
