This post is the first in a series, a conversation about the NY Times articles on the data center industry.

On Sunday, September 22 the New York Times published the first article in a series about the physical structures that make up the cloud, and their impact on our environment. The first article, “Power, Pollution and the Internet,” focused on the inefficient operating practices rampant in the data center industry. Kenneth Brill, Hank Seader and Bruce Taylor of the Uptime Institute were quoted in the article. The Sunday issue of The New York Times has a circulation of just over 2 million and NYTimes.com is the most popular American newspaper website averaging more than 30 million hits each month.

Uptime Institute is pleased to see this issue receiving attention in the national media, but the initial article on the front page of the Sunday New York Times did not recognize the progress that’s been made by the industry, and did not reflect the path that the industry is on for even further efficiency gains.

In fact, in the six years since Uptime Institute took up the task to improve the economics and sustainability of global enterprise IT, the data center industry has dramatically improved energy efficiency.

These improvements have come from the data center facility management and design field – the traditional constituency of the Uptime Institute – and are largely attributable to data center designers and operators providing more efficient cooling of IT equipment.

Companies are taking a number of steps to improve data center efficiency, especially related to airflow containment, increased inlet air temperature on servers, and increased monitoring of cooling. This is a positive trend in terms of cost, efficacy, and enhanced consciousness of the industry overall. These improvements can typically be accomplished inexpensively and with current staff resources.

The graphic below from the Uptime Institute 2012 Data Center Industry Survey outlines the steps large organizations (data centers managing over 2,000 servers) are taking to improve data center efficiency. These large organizations have the resources and financial incentives to wring the most effectiveness out of their data center infrastructures.

Let’s try to put this efficiency improvement into quantifiable terms, specifically improved PUE.

Power usage effectiveness (PUE) is a metric used to determine the energy efficiency of a data center’s facility infrastructure. In 2007, Uptime Institute surveyed its Network members (a user group of large, traditional enterprise data center owners and operators) and determined the average enterprise data center had a PUE of 2.5. This means that for every 2.5 watts in at the utility meter, only one watt is delivered out to the computing load. According to our 2012 survey, featuring 1,100 enterprise data center end users, the average global, self-reported PUE is between 1.8-1.89, a significant improvement over early estimates.

Uptime Institute acknowledges that self-reported, averaged PUE numbers are a blunt tool for assessing data center infrastructure efficiency. But companies measure PUE with increasing accuracy and detail and PUE is an accepted indicator of the industry’s progress.

Steps to further improve mechanical efficiency are necessary but will be incremental, and minor compared to the potential for energy savings on the IT side, as reducing the IT load has the double benefit of reducing the required mechanical load as well. Yet, IT practitioners are not incentivized to reduce energy consumption, as typically IT organizations are not responsible for paying the power bill.

Only 20% of IT Departments pay the power bill

The first round of data center efficiency gains were low-cost improvements, self-funding projects. Facilities executives led the first charge to improve data center energy efficiency because the cost of inefficiency was allocated to their department. Uptime Institute has long maintained that future improvements in data center efficiency will depend on incentivizing IT practitioners to take the next steps.

IT operations staff can drive exponential improvements in data center efficiency and effectiveness. IT organizations that will benefit most are those that take a systematic approach, starting at the application and data layers: consolidating applications and servers, de-duplicating data, removing comatose but power-draining servers, building redundancy into the applications and IT architecture rather than physical systems, improving server utilization; these initiatives will result in the most significant efficiency gains and drive the next wave of energy-efficiency innovation. But today, according to the 2012 survey data, only 20% of organizations’ IT departments pay the data center power bill.

Driving behavior and deployment changes to IT from the facilities operations side is a difficult task, but there is one area where both sides of IT delivery can come together and make a significant impact on data center energy consumption is killing comatose servers.

Uptime Institute conservatively estimated up to 10% of enterprise servers are running obsolete or unused software, have no function at all, yet remain in operation. Decommissioning a single 1U rack server can result in $500 per year in energy savings, an additional $500 in operating system licenses, and $1,500 in hardware maintenance costs. But it takes hard work to identify comatose servers and many organizations do not go through the effort. Last year, Uptime Institute introduced an annual contest (Server Roundup) to encourage the removal and recycling of obsolete IT equipment in an effort to decrease data center energy use.

Uptime Institute strongly encourages participation in this year’s contest, details and rules posted here. Last year’s first place winner, AOL, rounded up close to 10,000 servers and saved its organization over $5 million.

Since 2006, the mission of Uptime Institute Symposium has been to continuously increase uptime and global IT productivity through benchmarking and collaborative learning. Over the years, the Institute focused on identifying practical, immediate data center design and operations improvements that yield significant energy savings without capital expenditure, and offering strategies for breaking down the organizational barriers that undermine progress.

Uptime Institute Symposium has been a driving force for progress in Green Enterprise IT – challenging the largest data center operators in the world to deal with runaway energy consumption, making IT energy efficiency a C-level issue, and providing the platform that launched The Green Grid.

Engage with us at our annual spring event that brings together industry stakeholders in enterprise IT, finance, executive management, data center facilities, and corporate real estate to deal with the critical issues surrounding enterprise computing. At Symposium, the best minds in the business meet to share knowledge and discuss how the IT industry can meet its collective goals of performance, availability, energy efficiency, and profitability.

At Symposium, Uptime Institute grants Green Enterprise IT (GEIT) Awards to projects, ideas, and products that significantly improve energy productivity and resource use in IT. The Awards are open to applicants in all countries and are carefully judged by a committee of independent experts. These provide groundbreaking, repeatable case studies and the industry’s highest honor for corporate IT sustainability.

Resources:
-Download the full survey report
-Sign up for the Server Roundup
-Read about how AOL saved over $5 million removing 10,000 obsolete servers
-Stay up to date on Uptime Institute Symposium
-Apply for the 2013 GEIT Awards

Response to the Inaugural Uptime Institute Server Roundup Contest has been great so far, with teams signing up from around the world to participate. The goal of the event is to remove obsolete servers, save energy, and save money. Decommissioning a single 1U rack server can result in $500 per year in energy savings, an additional $500 in operating system licenses, and $1,500 in hardware maintenance costs. That’s not chump change.

Winners of the contest will receive one of these beautiful rodeo belt buckles, just finished by cowboy artisans in Texas:

Winners will also be honored at Uptime Institute Symposium 2012 and their case studies will be featured sessions.

It’s not too late to sign up! Click here for the rules, and contact Uptime Institute’s Hank Seader with any questions. Deadline for the contest is February 1st, 2012.

Uptime Institute recently participated in SearchDataCenter.com’s Data Center Advisory Board Q&A on capacity planning. In the article, Steve Carter, Uptime Institute VP of Digital Infrastructure Services explains how companies can start looking at long-term planning.

Carter warns that traditional capacity planning forecasting practices will lead to over-building:

“Many new data centers are significantly over built. My belief is that using common capacity planning forecasting practices in use today will lead to at least 2X over specification of spatial requirements. Common capacity planning practices are not keeping pace with the realities of advances in hardware and the software that enables visualization/consolidation capabilities.

“I believe this gap is and will continue to lead to over-specification of total data center infrastructure requirements from the IT load perspective. Understanding how application load drives IT infrastructure requirements, and how IT infrastructure requirements drive data center facilities requirements needs to be better understood at more granular level. In the past, a lot of this was guesswork. We can no longer estimate future IT electrical loads by simply projecting past IT load parameters based on historical UPS electrical load profiles.”

Family owned Central-Oregon cable company BendBroadband only recently decided to get into the data center business, but in that short time the organization has quickly proven it wants to be a leader in the industry.

The company earned Uptime Institute Tier III Facility Certification, and was the first site to be certified with a Kyoto Cooling system. In fact, the BendBroadband Vault is one of the biggest Kyoto Cooling installations in North America.

BendBroadband was also awarded the U.S. Green Building Council’s LEED Gold status, under the LEED 2009 Building Design and Construction (LEED BD+C) rating system. From the BendBroadband blog: This is a monumental achievement and one that we have been working toward since the initial concept stage of the Vault. This certification puts us in the upper echelon of data centers and makes us only the 5th data center in the world to attain this level.

BendBroadband is also an early adopter of Data Center Infrastructure Management (DCIM) software. The company recently announced it is using nlyte software for capacity planning.

I spoke with Steven Hall, Data Center Director at BendBroadband about the company’s DCIM use. Hall uses nlyte to help the organization bring new customers into the data center. Clients provide BendBroadband with a list of the servers they want to deploy, and Hall plugs those models into nlyte to plan out how much space, power and cooling will be necessary to meet the workload, and offers various options for deploying the equipment.

“We use it as part of our on-boarding process,” Hall said. “We help customers take a quick look at different cabinet layouts. Do they need a high density cabinet, or should we spread it out over two cabinets? The tool was perfect for that.”

Uptime Institute Executive Director Pitt Turner wrote a column this month for Mission Critical Magazine, exploring the top challenges facing data center executives over the next five years. Turner describes the future of data center financial management, availability requirements, operational best practices, and staffing.

Read the piece here.