The hottest data center technologies and trends
From liquid cooling to lithium batteries, we take a look at some of the hottest technologies and trends that will change the way data centers are built or managed.
Even as the data center industry continues to grow fast on the back of huge demand, the industry is known for its constant innovation. For example, major data center players are using technology innovatively to cut down their power consumption. A recent report by the Lawrence Berkeley National Laboratory states that improved energy is almost canceling out growing capacity.
Looking at the rapidly changing landscape, we believe that the following key technologies will play a big role in shaping the future of data centers:
1. Liquid cooling
With power costs being one of the major cost areas for a data center, optimizing data center power use is a high priority area for data center managers. For high power density servers, Gartner estimates that ongoing power costs are increasing at least 10% per year due to cost per kilowatt-hour (kwh) increases and underlying demand. Typically, Gartner estimates that 10% of data center OPEX is power, and this number will only increase. The research firm also believes that a major percentage of power consumption (close to 40%) is attributed to cooling methods.
To reduce energy related costs, liquid cooling has emerged as a powerful and popular technology. As liquids have superior thermal efficiency, they are able to cool down the equipment directly at the source of heat. This can be done by submerging the equipment in liquid or by implementing direct-to-chip cooling.
A report by Research and Markets states that liquid cooling technology is gaining popularity among high-computing facilities to support Artificial Intelligence (AI) and big data analytics. These technologies require powerful computing capability for their processors where the heat generated exceeds the cooling limit, and thus they need efficient cooling to keep the chips functional.
2. Lithium-ion batteries
Adoption of lithium-ion batteries is rising fast in data centers, as they pose a good economic alternative to lead batteries that power UPS systems. A report from Research and Markets predicts that lithium-ion batteries will account for 35% of the market share for UPS batteries by 2025. While the initial costs may be higher than a conventional lead battery, lithium-ion batteries are being preferred as they can last conservatively about 2-3 times more than a lead battery. A report from Research and Markets says that organizations can easily end up saving more than 30-50%. With a smaller footprint (roughly 40-60% less size), data center owners save lot of space. A typical li-ion battery can also be fully recharged in approximately two hours, unlike a lead battery that requires 10 hours. Due to these advantages, research firms like Bloomberg New Energy Finance, predict that lithium-ion batteries will quickly command a growing portion of the data center UPS market.
3. Direct cloud connections
Latency is one of the biggest challenges for an organization in the hyper-connected world. Users will quickly leave a website if it fails to load in a matter of seconds. Leading global cloud providers have realized this trend, and have hence created 'direct cloud connections'. This gives organizations a direct connection between their on-premise environments to the public cloud. Each one of the major cloud providers has this option. Amazon has the Direct Connect, while Microsoft offers ExpressRoute and Google offers the Dedicated Interconnect. As these bypass the internet, they do not have to compete for internet bandwidth and also reduce exposure to unsecured public networks. Typically, direct cloud connections offer higher reliability, security and speeds with lower latencies.
AI can play a big role in automating many of the tasks that human agents do today. In the future, expect AI to handle every component or function of the data center – be it network design, managing and predicting IT workloads or saving energy. This is critical in today's age where network administrators are often overwhelmed with the large amount of data gathered through multiple channels. Gartner, for instance, predicts that by 2020, 75% of organizations will experience visible business disruptions due to I&O skills gaps (an increase from less than 20% in 2016).
Today, AI can even complement current security incidents and event management (SIEM) systems, by analyzing incidents and inputs from multiple systems, and devising an appropriate incident response system. Naturally, analysts are bullish on the impact of AI in the data center. By 2022, IDC predicts that 50% of IT assets in data centers will be able to run autonomously because of embedded AI functionality. Further, Gartner has predicted that more than 30% of data centers that fail to sufficiently prepare for AI will no longer be operationally or economically viable by 2020.
About the author
Vimal Kaw is currently the Associate Vice President, Products and Services at Netmagic. In this capacity, he is responsible for supporting the technical aspects of co-location business and providing end-to-end support for the entire value chain – from conceptualization and development of collocation products and services, to regular pre-sales customer engagement, solution / service design and customer project delivery.