Issue link: https://www.e-digitaleditions.com/i/1107553
May/June 2019 I 23 Change may seem justifiably overwhelming for data center operators, engineers, designers, consultants, project managers, and installers who must simultaneously attend to their daily responsibilities while keeping up with the large and continuous influx of new technologies and paradigms. The new ANSI/BICSI 002-2019, Data Center Design and Implementation Best Practices standard provides assistance and a path forward amid ongoing changes and trends in the age of the Internet of Things (IoT), big data and ICT transformation. A CLOUD IS BORN In 2009, enterprise data centers were seemingly all the rage. Store the data internally to the company, and provide it to staff as needed. To most people, reference to "the cloud" meant something that existed on the horizon that could bring rain; those in data center circles thought similarly. Except to them, the cloud was bringing a new paradigm. Shortly thereafter, the cloud entered common use. The cloud became an easy to use metaphor to describe data that was stored somewhere and waiting to rain down on the user. Use of the cloud grew quickly as some companies saw financial benefits for having "someone else" store information that reduced internal infrastructure and labor costs. Other companies viewed the cloud as the provider of much needed redundancy for critical data systems. As demand grew, so too did the FIGURE 1: Data traffic is similar to vehicle traffic; both need available pathways and space. number and physical size of data centers. Data center construction was in a steady growth pattern as more applications drove people and certain areas of the globe to create and demand more data. As construction was trending upward, technological advancements were performing as historically expected by decreasing features and function sizes of key items and allowing for more powerful servers, storage systems and network equipment to reside in the same physical space. Because the ability to send data from point A to point B depends on the pathway and space available, not unlike automobile traffic (Figure 1), advancement in networking protocols allowed networks to start determining optimum routing. These advancements and growth occurred in the course of about eight years and led to the realm of the hyperscale data center. While the term "hyperscale" often connotes a very large computer room area in excess of 9,000 m 2 (100,000 ft 2 ), hyperscale may also refer to a data center with as little as 5,000 servers and 929 m 2 (10,000 ft 2 ) that can provide high-volume traffic and the ability to handle heavy computing workloads for organizations that run most of their applications in the cloud. 1 Granted, having additional space assists with the potential volume and proper selection of equipment and components to fit specific use case requirements, but running a data center is much more than about space. The data center, regardless of size, must also be operationally efficient.