The next time you look to your smartphone to recall the name of that actress who co-starred in that movie with the guy from that television series, consider a study which estimated that a typical internet search engine uses as much energy as illuminating a 60-W lightbulb for 17 seconds while emitting 0.2 grams of CO2. While that might not seem like a lot, consider next how many web searches you conduct in a day, a week, a year. Now multiply that by the 4.33 billion people across our planet that Statista reports were active internet users as of the summer of this year.

On a mass scale, the energy footprint of the IT sector is estimated to comprise 7% of global electricity consumption, according to Greenpeace Intl. The data center industry, one small segment of the IT business, represents a large portion of its power consumption, drawing approximately 3% of all power generated globally and accounting for roughly 2% of greenhouse gas emissions.

To bring these percentages into even starker relief, five years ago, a study by the U.S. Department of Energy (DOE) found that the nation’s data centers consumed roughly 70 billion kWh of electricity, an amount equal to that consumed by about 6.4 million American homes. Globally, data centers consume approximately 30 billion watts of electricity, an output equal to 30 nuclear power plants, or enough to keep the lights on in all the households of Italy.

Yes, data centers are the digital factories of our connected age — and whether you are sending a personal email, using a web-based collaboration software system at work, or binge-watching Netflix — moving, storing, processing, or analyzing all that data requires energy … and lots of it. Cisco estimates that bandwidth-intensive video will make up 82% of all internet traffic by 2021, up from 73% in 2016. And then, of course, there is the surge in data that will accompany the 50 billion IoT devices projected to come online by next year, twice the number in 2015. Overall, IDC predicts the world’s data will grow 61%, from 33 zettabytes (ZB) in 2018 to 175 ZB by 2025.

While we cannot reduce the amount of data that emerging and next-generation technologies will generate, the good news is there are simple choices and innovative solutions that can be employed in and around the data center to reconcile the energy demands of our digital services with the environmental mandates of our planet.


Not Every Data Center Can Border the Arctic Circle

In the effort to make their data center operations more energy-efficient and sustainable while reducing their carbon footprint, many of the world’s tech giants have turned to renewable energy sources. Since 2014, all of Apple’s data centers have been powered by 100% renewable energy. Two years ago, Google announced that all its data centers were operating on 100% renewable electricity. Amazon Web Services exceeded 50% renewable energy for its global infrastructure in 2018, and Microsoft reached 50% renewable energy last year and plans to shift its data centers to 75% renewables by 2023.

According to a 2018 study DigiPlex conducted in partnership with IDG Connect, sustainability has emerged as the fastest rising data center concern among C-levels, rising to fourth place from 27th in just two years, and joining operating costs, energy, and security as among the most important. A number of data center builds have recently taken place “up north,” from 100 or so kilometers shy of the Arctic Circle in Sweden, to Finland and Norway, and to Québec, Canada. In addition to abundant and affordable renewable energy resources, these locations offer relatively long winters; shortened summers; and in one instance, fjord-chilled waters, which make it easier to maintain proper server temperatures, lower cooling costs, and reduce OpEx.

There’s no question that all these renewable energy resource and climate-driven data center locales make for sound environmental stewardship and good business. But the reality is that not every future data center can be located near the Arctic Circle. Today’s distributed data environments often comprise on-premise, public, private, and hybrid cloud, as well as colocation facilities spread across a single enterprise with multiple, geographically dispersed locations.

Fortunately, data center management tools can be quickly implemented on a large scale for a low cost, resulting in a decrease in IT expenditures and a reduced environmental impact.


Stanley, How Can We Still Know Ye?

In an era of automation, it seems counter-intuitive that a recent study found one in five data center managers still rely on manual processes, tape measures, and spreadsheets to perform jobs that could be made more efficient through innovative software solutions.

Converging IT and building functions to provide a 360-degree, holistic view of a facility’s performance, data center management tools ensure that energy, equipment, and floor space are used as efficiently as possible.

In large data centers, where electrical billing comprises a significant portion of the cost of operation, real-time insights into power and thermal management have a direct and positive impact on an organization’s bottom line. In fact, a study by Gartner estimated that data center management tools can reduce operating expenses for companies by as much as 20% or more.

Accurate, real-time power and thermal monitoring of individual servers, groups of servers, racks, power distribution units (PDUs), and other IT equipment enables collaboration among IT and facility administrators to reduce a data center’s energy footprint, increase efficiency, and improve uptime.

This last value cannot be overemphasized. Make no mistake, IT executives that manage tens of thousands of data assets will experience their fair share of system failures or problems. Although figures from various studies vary greatly, one conservative report, the “Global Server Reliability Survey” conducted by the Information Technology Intelligence Consulting (ITIC) group, found that 50% of large enterprise data centers experience downtime at an average cost of $300,000 per hour. To offset the potential of these vulnerabilities, data center management software can be utilized to project and forecast peaks in server capacity and usage, thereby minimizing unexpected downtime.

Ongoing monitoring, analytics, diagnostics, and remediation, with granular subcomponent failure analysis across racks and servers accessible from a single pane of glass, can identify issues before they occur. Additionally, a survey by Redshift Research and commissioned by Intel® found that 21% of data center managers using these tools for capacity planning and forecasting stated that they could recover from an outage within two hours, compared to only 11% of those without the software.


Under Pressure but Keeping Cool

By some estimates, 30 to 60 percent of a typical data center utility bill goes to cooling systems. However, many data center administrators overcool their systems, wasting energy in the process. Data center management software solutions aggregate server-inlet temperature data into thermal maps, which allows IT staff to monitor the effectiveness of their cooling solutions and airflow design.

By providing advanced thermal management capabilities, data center management tools can target which areas of the data center are overcooled and determine those that are vulnerable to damage in the event that temperatures should rise, for example, during a heat wave. Ultimately, this critical preemptive data allows data centers to operate at higher temperatures while lowering their cooling costs. For the average 300-rack, 3-MW facility, a 4°F temperature increase can save 20 percent in cooling costs.

Data center management solutions also identify underused servers. Servers use 50% of a data center’s power, even when idle. In large data centers, a lack of sufficient workload performance monitoring typically leads IT administrators to purchase more hardware. Using data center management software, data center operators are able to quickly detect and analyze underutilized systems by monitoring their CPU utilization and power consumption over time. In addition to saving power, this helps to identify opportunities for consolidation and virtualization.

According to one government environmental watchdog, data centers consume as much as 20 times more energy per square foot than a typical office building. To accomplish meaningful reductions in their carbon footprint and operating expenditures, no matter where their facilities are located, data center operators must first be able measure energy-related data across their entire site, including the building, the facility’s components, and the IT equipment portfolio.

Data center management tools, by maintaining visibility on power, thermal consumption, server health, and utilization, enable better data-driven decision-making and more precise operational control. A solution that is good for the data center is good for the planet, and that’s simply good business.