<img src="https://d5nxst8fruw4z.cloudfront.net/atrk.gif?account=Wfzuj1agkg00y6" style="display:none" height="1" width="1" alt="">

Why is a Data Center Like a Supply Chain?

Posted by Elementum News Desk | March 16, 2018

They don’t seem too similar, but data centers could benefit from supply chain management.

Last week Facebook revealed that it will be building a new data center — its ninth in the US — in Atlanta, Georgia. The $750 million investment will run completely on renewable energy (most likely solar). It’s not hard to imagine how much power the information of 2.2 billion users might require, and renewable options cut costs in the long-term.

But that’s just one example. In 2014, there were an estimated 3 million data centers in the United States alone. Although new centers are expected to trail off in a few years, that doesn’t mean there will be less data to process. In fact, existing data centers will be expanded upon to create mega data centers (over 225,000 square feet of computer space), which will account for more than 70 percent of all center construction by 2018, according to Data Center Knowledge. Regardless of whether companies need new or expanded centers, it remains true that these centers use a lot of power, money, and natural resources. Communication and privacy worries also abound, and in such large premises, it’s easy to lose track of issues.

Given data's diverse use of resources and the fragmentation that can occur when operations are scaled, data centers are similar to supply chains. Supply chain management (SCM) technologies can thus prove incredibly useful, from location scouting to communication & logistics.


The Build-Up

Data center construction is centered around choosing the best location, which encompasses regional and national codes and laws, risk factors, and privacy regulations. Different countries require different considerations, which, for multinational corporations, can be difficult to navigate. When building in the United States, the cost of land and various state taxes and fees should be taken into account. Interestingly, a lot of data centers are located in or near busy hubs and often in disaster-prone states. There’s no need for data centers to be in industrialized areas, according to FlashGlobal, other than for access to staff and construction workers.

Land considerations like this are a perfect job for SCM. Supply chains are essentially all about location, and the best management systems include an analysis of potential sites, weighing factors like cost, local policies, and risk, to find the most cost-effective location.

Once a location has been secured, companies must consider the costs involved in construction and maintenance of the data center. Planning usually consumes 20-25% of the total land cost — buying land, obtaining permits, and lawyer fees all add up. After construction, the power required to maintain the center eats up most of the remaining 70-80% of the total budget.

With so much data to store and transfer, these centers carry both a huge energy expense and a huge risk when it comes to power outages or natural disasters that could compromise data. Indeed, power is where many companies can utilize SCM techniques to cut costs and turn to more sustainable techniques. Risk management and contingency plans are both common supply chain areas that should be utilized in a data center.


The Punchline

In addition to location, SCM focuses on building effective communication between supply chain “nodes”. Data centers, when viewed like entire supply chains, can be divided into specific roles that function as nodes. But often, those nodes’ ability to communicate quickly and efficiently is overlooked, leading to wasted time and money. According to Data Center Knowledge, similar problems often arise in multiple departments. With no clear communication channels, there is no way to link the problems and come up with a universal solution. Similarly, if one department comes up with a way to streamline an activity, other departments should be made aware. As great supply chains have shown, communication and collaboration are keys to smooth operations.


Final Thoughts

Data centers are an increasingly necessary part of the modern world, but they don’t run cheap. And while their size may be intimidating, breaking various functions into nodes — just like a supply chain — can help companies figure out more efficient, cost-effective construction and maintenance techniques. There are worlds of data inside every center — SCM techniques will make navigating them that much easier.