The Story so far
Since the mid-20th century, data centers have been intrinsic to the growth capabilities of businesses. While in the beginning, data centers served very specific purposes such as a processing system to manage airline ticketing, with the advent of computing technology, mainframes came into existence along with a significant enhancement in the role a data center plays in an enterprise’s scheme of things. The data center-as-a-service model was born of the requirements for enterprise-wide faster connectivity and uninterrupted operations.
Some of the leading technology firms have, since the turn of the century, been busy exploring cross-platform operability, modularity and cloud computing. Data centers across the globe are in the middle of a shift toward a subscription and on-demand capacity model. In order to support the rising data demands, the data center industry is undergoing a trend of consolidation and cost efficiency powered by the cloud.
Given the volume of data that is already defining businesses globally, data centers will play an essential role in the collection, storage and consumption of information in the times to come.
Today’s data centers find themselves in close proximity to dense fiber networks. Consequently, some of the bigger technology firms are toying with the idea of building underwater transcontinental cabling.
3 Key factors driving the Data Center evolution in the digital age
- Energy imperative
Data centers are energy-hungry establishments and organizations are already constructing them near cheaper power sources. Another concern is the availability of clean energy to run data center operations. In light of this, data centers are being built in the Nordic regions which offer natural cooling, thanks to the climate of the region. The likes of Microsoft are experimenting with the concept of submerged data centers which allows the submerged entity to have unrestricted access to the naturally cool sea water to ensure better energy efficiencies of the entity. Google has been experimenting with the idea of applying AI and deep learning to induce adaptive energy consumption based on traffic trends observed toward its servers.
- Space imperative
Data centers, being at the center of technological transformation, are also being transformed with respect to their size. As cloud computing soldiers forth toward industry-wide adoption, various alternative computing techniques such as edge computing and modular computing are being explored to serve very specific needs. With edge computing, the idea is to place computing resources close to the origin of data. The case of autonomous vehicles presents an excellent use case as having powerful on-board computing enables near real-time feeds from the AVs which can then be used by the vehicle to make timely and often life-saving decisions.
Modular data centers, on the other hand, help serve very specific business needs by offering a targeted computing capacity. A case in point could be a mobile network operator that wishes to deliver a particular kind of content in a geographical area. A modular data center would serve as a compact solution to ensure efficient storage and faster delivery of content to intended users.
- Storage imperative
A key function of every data center lies with its storage media. Having moved from magnetic tapes to CDs to HDDs, solid state drives are something to watch out for in the next 3 to 5 years as they provide a more responsive and more durable form of mass storage. SSDs have already begun making a strong case for rivalling the HDDs as the preferred storage medium. Once their manufacturing process and costs are optimized, it is only a matter of time before they become widely accepted. On another front, Seagate has been working on the laser-assisted HAMR (Heat assisted magnetic recording) technology which holds a promise to ensure faster and more accurate data recording.
The road ahead
Looking forward, we should expect to see the data center landscape characterized by the ability to add on-demand capacity rapidly, adopt innovative technology (Eg. P2P device networks), share computational resources, and utilize HAMR, etc. to provide storage at lower costs.