Page 204 -

10: Green IT Case Studies for Universities and a Large Company 169

Columbia University Green IT Plan Summary

The Columbia University plan for green IT will provide valuable input for
IT energy-efficiency projects for all organizations:

■ The project goal of leveraging current best practices and latest lessons
learned from all the organizations involved in green IT fits with the key
need to have wide collaboration across many organizations so that every-
one can benefit. In addition, the plan to communicate lessons learned to
peer higher education institutions and other groups will help us all.
The challenge of the green computing and the increasing demands of
HPC might be unique to research universities and research compa-
nies. However, concepts used to meet the requirements of these com-
putationally intensive systems in an environmentally friendly way
will prove to be valuable to all IT teams.
■ The plan includes testing of power and cooling efficiency techniques via a
pilot data center before deciding which techniques would be best for
Columbia’s central data center. This will provide valuable lessons learned
ptg
for all data center teams. Establishing a baseline and continuous measure-
ment of variables is fundamental to success and should be a best practice
for all groups.

Lexington, Kentucky, Large (61KSF) Data Center

As data centers are becoming more constrained, it is essential that capacity
managers step to the forefront and help out. This case study analyzes one
such very constrained, very large data center. It describes a five-step approach
to mitigating data center constraints, delaying a costly data center upgrade
and data center relocation. Best practices will be described, including data
center level metrics in the industry and recent governmental work in the data
center metrics. The case study is based on information provided by Chris
Molloy, an IBM Distinguished Engineer from Raleigh, North Carolina, who
led the IT team on this green data center project.
In an IT environment of on-demand computing with virtualization getting
a significant amount of attention, it is easy to forget that IT equipment