We are living in a lucrative period for technology providers, but with the technological tsunami sweeping the market, end customers are increasingly overwhelmed by it all. Artificial intelligence to the rescue? Hold that thought.
That’s the word of a recent Uptime Institute survey that shows that technology providers and data center operators are growing almost continuously. For the third year in a row, around 80% report that customer spending is at or above normal levels. strong growth in capital spending over the next three to five years. About one in three expects growth rates to slow, but only one in six expects them to flatten or shrink.
Good news, but where is all of this new stuff going? Predicting data center capacity remains a dark art, and for the fourth straight year, vendors say it’s the biggest problem for their customers. Managing a combination of different data center environments — typically any mix of on-premises, co-location, both in retail and wholesale, and the cloud became the second biggest challenge for the operator.
Additionally, nearly half (47%) of data center operators say they have difficulty finding qualified candidates, and 38% of technology vendors say staff shortages will stifle growth. While AI is expected to become more widespread in the next five years, it will not alleviate the staff shortage. One approach is to open up the hiring to a wider variety of potential candidates, with the majority of vendors (88%) expecting even more diverse employees to be hired in the next three to five years.
Both technology providers and data center operators share the view that AI is increasingly being used to improve facility operations. Artificial intelligence techniques are advancing and the pandemic has forced many data center operators to review their investments in remote monitoring and related software, which has sparked interest in the technology, said survey authors Rhonda Ascierto and Jacqueline Davis of the Uptime Institute State. The growing proportion of vendors (89% in 2021 (up from 70% in 2019) agree that AI will be widely used in data centers over the next five years to improve efficiency and availability.
Only about a third believe that artificial intelligence will reduce data center employment in relation to workload over the same period. Only 23% of operators expect the technology to reduce the number of operating employees in the next five years. Operators take a measured approach to deploying AI-powered software in business-critical functions, say Ascierto and Davis. “Improvements in technology and its ability to reduce headcount through automation are often the result of an iterative, human-driven process. Building trust in a system takes time.
Uptime analysts predict that by 2025, nearly 2.3 million full-time employees will be needed to support the design, construction and operation of data center infrastructures worldwide, compared to an estimated two million in 2020. Many existing workers, the Authors state.
The biggest challenge for customers is consistency: anticipating the capacity of the data center. For some operators this means managing the uncontrolled demand in their own data centers on site, while others expect demand to decrease as more Work is being shifted to third party infrastructure, according to Ascierto and Davis. It’s about where different workloads should run based on cost, resilience, compliance and other requirements.
Edge computing has also made its way into technology design and capacity planning. one year ago. Vendors’ confidence in instant edge growth has increased. 60% agreed that the majority of their customers will own small data centers at the edge within five years, up from 48% a year ago. Most, almost three-quarters, say they make changes to their products or services in response to the marginal opportunity. “Redesigns and modifications may be required for edge deployments because edge computing is often located in remote or new locations with little or no staff presence,” state the authors of Uptime.
Over the past decade, availability Zones have become the de facto standard for hyperscale operators to keep service running, but the focus is no longer limited to hyperscales, they add. Organizations are starting to roll out private cloud workloads in racks in at least three colocation locations that are close enough to ensure low latency but far enough apart to avoid localized failure in a location that is affects the workloads on another. By replicating and redirecting workloads between sites, this means that the need for fault tolerance in a single data center can be reduced, at least in theory.
There are two reasons for this diversification: inceased independence from IT services and architectural complexity. “The data center is more important than ever, and critical component failures can cascade, making recovery difficult and expensive,” say Ascierto and Davis.
Awareness of the role IT plays in sustainability is growing, as the survey shows. The expected construction of more large data centers (20 MW or more) over the next five years by suppliers will put pressure on the industry to improve. Almost three quarters of suppliers expect most facilities to have CO2 reduction targets by 2022; Almost four out of five suppliers have set their own goals to reduce the environmental impact of their products or services. Most of these suppliers (80%) have revised their data on center products / services in order to improve their ecological footprint in the past two years.
