This is a little note to advise universities working on Hadoop related
projects, that they may
be able to get some money and cluster time for some fun things
http://www.hpl.hp.com/open_innovation/irp/
"The HP Labs Innovation Research Program is designed to create
opportunities -- at colleges,
universities and research institutes around the world -- for
breakthrough collaborative research
with HP.
HP Labs is proud to announce the 2009 Innovation Research Program (IRP).
Through this open Call for
Proposals, we are soliciting your best ideas on a range of topics with
the goal of establishing new
research collaborations. Proposals will be invited against targeted IRP
Research Topics, and will be
accepted via an online submission tool. They will be reviewed by HP Labs
scientists for alignment
with the selected research topic and impact of the proposed research.
Awards under the 2009 HP Labs Innovation Research Program are primarily
intended to provide
financial support for a graduate student to assist the Principal
Investigator conducting a
collaborative research project with HP Labs. Consequently, awards will
provide cash support for one
year in the range of USD $50,000 to $75,000, including any overhead."
If you look at the research topics there is a PDF file listing topics of
interest, of which three
general categories may be of interest:
http://www.hpl.hp.com/open_innovation/irp/topics_2009.html
1. "Intelligent Infrastructure" - very large scale storage systems,
management, etc.
2. Sustainability -especially sustainable datacentres: how to measure
application power consumption,
and improve it; how to include knowledge of the physical
infrastructure in computation
3. "Cloud" - Large-scale computing frameworks, Data management and
security, Federation of
heterogeneous cloud sites, Programming tools and mash-ups, Complex
event processing and
management, Massive-Scale Data Analytics, Cloud monitoring and
management.
If you look at that Cloud topic, not only does Hadoop-related work seem
to fit in, the call for
proposals is fairly explicit in mentioning the ecosystems suitability as
a platform for your work.
Which makes sense, as it is the only very-large-scale data-centric
computing platform out there for
which the source code is freely available. Yet also, because it is open
source, it is a place where,
university permitting, your research can be contributed back to the
community, and used by grateful
users the world over.
What is also interesting is that little line at the bottom, "We
encourage investigators to utilize
the capabilities in the Open Cirrus testbed as well as to share their
experience, data, and
algorithms with other researchers using the testbed. " Which implies
that cluster time on the new cross-company, cross-university homogenous
datacentre test bed should be available to test your ideas.
If you are at university, have a look at the proposals and see if you
can come up with a proposal
for innovative work in this area. The timescales are fairly aggressive
-that is to ensure that
proposers will know early on whether or not they were successful, and
the money will be in their
University's hands for the next academic year.
-Steve
(for followup queries, follow the links on the site or email me direct;
I am vaguely involved in some of this)
