Applied Research
Connect with AIM
UO AIM Program on Facebook UO AIM Program on Twitter UO AIM Program on LinkedIn UO AIM Program RSS feeds UO AIM Program on YouTube Contact the UO AIM Program
Do you have a question or comment? We would love to hear from you. Get in touch.
Pictured Above
Amy R. Cissell ('12) receives the 2012 Capstone Award from Dr. Linda Ettinger for her research paper, Examining the Role of Information and Communication Technologies to Improve Food Security Management: The Case in Oregon at the AIM graduation luncheon.

Lowering the Carbon Emissions Footprint of Enterprise Data Centers Through Energy Efficiency Gains

In Brief: In 2007, Gartner found that information communications and technology is responsible for approximately two percent of global carbon dioxide (CO2) emissions, which is as much as the aviation industry produces. Gartner also believes this rate of emission is unsustainable. Furthermore, Gartner declared that the energy consumed by data centers is responsible for almost a quarter of global carbon dioxide emissions produced by information and communications technology.

The purpose of this study is to develop a set of recommended modifications that, when applied within enterprise data centers, may achieve lower carbon emissions. The focus is on specific changes that are practical to implement within existing facilities.

There is an urgent need to define metrics for IT performance, scalable to the data center.

The data center supports information technology and communications, which in turn supports information management. The effort to decrease energy consumption and the overall carbon footprint of data centers should be a goal for the industry as a whole. Many organizations are looking for small incremental changes to help lower emission costs without a large capital investment and without disruption to daily business. In an effort to minimize expenditures, data centers are attempting to postpone replacement of hardware and facility infrastructure components.

According to recent surveys, companies are increasingly assessing the impact of their IT needs on the environment. Many have incorporated green criteria in their procurement processes. However, there are still a number of hurdles facing the IT industry, including: (a) difficulty in accurately measuring data center efficiencies, (b) a lack of clear standardized roadmaps from equipment vendors on efficiency gains in future produces and finally, and (c) a lack of industry best practices and tools for managing power consumption.

As noted in the selected literature reviewed for this study, there are several recommended first steps that enterprise data centers can take when investigating efficiencies and power consumptions (see figure 1). Not surprisingly, these steps align with the basic business strategy: plan, implement, measure, and correct.

Steps Actions
Gather efficiency metrics.
Enterprise data center managers know how much is spent overall on energy, but they generally do not know where it is being consumed. To manage consumption effectively, it must first be measured. Application of the Green Grid metrics is a good start.
Plan and design a path to implement best practices.
As government agencies are exploring ways to regulate energy use, green principles are receiving more attention within IT organizations. This involves examination of innovative technologies and solutions to meet energy efficiency goals.
Integrate this key strategy for server deployment.
Maximize the use of virtualization in the enterprise data center thoughtfully. Do not join the one-third of corporations already drowning in virtual server sprawl.
Apply metrics to optimize efficiencies.
New approaches for improving data center efficiency through improved workload management are offered. Decrease idle server consumption and improve air management. Monitor infrastructure efficiency, use trending for improved planning, and employ heat and capping analysis. Adopt existing change control processes to include efficiency goals.
Focus on heating and cooling efficiency.
Alternative methods for modeling and managing data center airflow and cooling are being explored. Adopt the latest cooling technologies and innovations such as water-cooling rack door mounts. Consider placing heavy computational workloads in data center locations that are easier to cool. Determine minimum effective ambient temperature.

Figure 1—Steps for investigating efficiencies and power consumptions at data centers.


Research Paper Author: Cecilia Tomory, PacifiCorp—2010 AIM Graduate

Abstract: Selected literature published between 2004 and 2009 is mined for emerging trends and best practices regarding green IT and data centers. Environmental issues are becoming a serious industry concern, spurred by local and global initiatives. Steps are identified for enterprise data center managers, in order to measure, manage, and improve power consumption efficiency within their facility. Data center energy efficiency gains save power, lower energy costs, and decrease overall corporate carbon footprint.

Prospective Students

Applied Research
Current Students

About AIM