Page 82 -

3: Collaboration Is Key for Green IT 47

Universities Collaborate

Universities are in a unique position to collaborate on green IT. The
case study on Columbia University in Chapter 10, “Green IT Case
Studies for Universities and a Large Company,” gives an excellent exam-
ple on how Columbia is collaborating on green IT—within the univer-
sity, with other universities, with New York State organizations, and
with New York City. Columbia’s Business School’s Green Club has
already indicated its enthusiasm to collaborate in the green IT exercise.
This can help lay the foundation for Columbia’s plans to submit the
green data center results to Educause, NYSERNet, NYSgrid, the Center
for IT Leadership, the Ivy Plus consortium, and the Common Solutions
Group as a real-world case study. Columbia University anticipates good
attendance at the Open House Workshop at the conclusion of the proj-
ect. The vice president and chief information officer, the assistant vice
president for Environmental Stewardship, and the E-Science Task Force
have all endorsed this proposal. The opportunity to rigorously measure
recommended best practices and technological innovations in a real- ptg
world environment, validated by the scrutiny incorporated from the
beginning via the three potential user groups, can have a far-reaching
impact within and beyond Columbia. The Columbia green IT collabora-
tion also includes New York City and Mayor Michael Bloomberg’s ten-
year plan for New York City on reducing carbon emissions by 30 percent
based on 2007 levels. Columbia University has committed to that 30
percent reduction even in the face of greatly increased growth in High
Performance Computing (HPC) requirements fueled by the worldwide
research community.
In the past several years, HPC has been growing at every research uni-
versity, government research laboratory, and high-tech industry in New
York State, nationally and internationally. HPC is a cornerstone of scien-
tific research disciplines, many of which had previously used little or no
computing resources. Researchers are now performing simulations, ana-
lyzing experimental data from sources such as the Large Hadron Collider
at CERN, genetic sequencers, scanning-tunneling electron microscopes,
econometric and population data, and so on. These applications have led
to an explosion of computing clusters now being deployed throughout
Columbia, as well as in peer research institutions and New York State’s
biotechnology, nanotechnology, financial, and other industries; this
increase frequently requires construction of new server rooms, putting