As the world grapples to handle a tsunami of data, data centers are also evolving quickly. Rapid growth in the number of smart connected devices and a huge rise in consumption of data is placing enormous amount of pressure on the underlying data center infrastructure. Data centers have become so complex, that is no longer possible for only human beings to manage this rising complexity, without impacting performance and efficiency levels. A technology like AI can help immensely in helping organizations improve the efficiency of their data centers in a significant manner.
Such has been the impact of AI on data centers, that last year, Gartner predicted that by 2020, more than 30 percent of data centers that fail to implement AI and Machine Learning will cease to be operationally and economically viable. Google highlighted one of the first instances of the possible potential of AI in the data center, when it published research that it used AI in the data center to improve the power efficiency of its data center. In a span of just 18 months, Google used its AI powered Google DeepMind system to bring about a 40% reduction in the amount of energy required for cooling, which is equivalent to a 15% reduction in overall PUE overheads. Since then, many firms have followed suit to explore the transformational potential of AI.
Handling workloads efficiently: AI can help organizations automate workload management in the most efficient way. With the usage of AI and machine learning, patterns can be detected to learn from past data and distribute workloads across peak periods more efficiently. They can also be used to better optimize disk utilization, server capacity and network bandwidth. This was demonstrated last year by a team of MIT researchers. An AI-based system developed by MIT researchers automatically “learned” how to schedule data-processing operations across thousands of servers — a task traditionally reserved for imprecise, human-designed algorithms. MIT researchers said that by doing so, they could help today’s power-hungry data centers run far more efficiently. The researchers said that compared to the best hand-written scheduling algorithms, the researchers’ system completes jobs about 20 to 30 percent faster, and twice as fast during high-traffic times. Additionally, the system learns how to compact workloads efficiently to leave little waste. The results indicate that the system could enable data centers to handle the same workload at higher speeds, using fewer resources.
Staffing: Hiring people with the right skill sets is a massive challenge in the digital era. Gartner, for instance, predicts that by 2020, 75% of organizations will experience visible business disruptions due to gaps in I&O skills (an increase from less than 20% in 2016). AI can play a big role in automating many of the tasks that human agents do today.
Energy Efficiency: As seen from the example of Google, AI based systems can play a huge role in better optimizing heating and cooling systems, which in turn can help in reducing electricity costs.. AI can also be used to determine how organizations can best utilize resources, such as the most efficient time for performing certain type of tasks. AI can also be used to help in creating the design for a more efficient data center, and in detecting applications or servers that are rarely used. It can also be used to detect power hungry applications or servers, and recommend ways to move specific workloads to more efficient ones.
Security: AI can be used with great impact in a security operations center in a data center. AI can complement current Security Incidents and Event Management (SIEM) systems, by analyzing incidents and inputs from multiple systems, and devising an appropriate incident response system. AI-based systems can improve the security operations centre monitoring and basic L1 jobs can be reduced. For example, when more than 20,000 events per second are logged, it becomes difficult for human beings to monitor these events. AI based systems can help in identifying the malicious traffic from the false positives and help data center administrators handle cyber security threats more efficiently.
Proactive management of hardware: AI systems can help organizations in proactively managing the health of their IT infrastructure such as storage, servers or networking equipment. For example, by aggregating logs of different equipment, AI can unearth the root cause of failures and also proactively identify precursors of degradation of equipment. Anomalies, if any, can be reported to address the probable cause of failure, before the equipment fails.
