The sector is booming and maturing globally — but
management challenges lie ahead.
The recently released Uptime Institute Intelligence Report looks beyond the obvious trends of 2022 — that the sector is thriving, building out to the edge and recovering and adapting to the pandemic — and identifies some potentially challenging issues. These include sustainability reporting and the difficulty of achieving zero-carbon emissions, the uneven impacts of processor innovations, supply chain problems and the
concentration risk of cloud computing.
The critical digital infrastructure sector continues to expand and attract enviable levels of new investment. The coming year (and beyond)
promises opportunities but also challenges that only some data centers will be likely to navigate successfully. This report predicts some of the
potential outcomes ahead, including achieving zero-carbon emissions, reducing dependency on cloud, benefiting from processor innovation and ensuring supply-chain resiliency.
Heading into 2022, the mood is — guardedly — more positive. Although the threat of COVID-19 variants persists, we can nevertheless repeat many of the sentiments from our Five data center trends for 2021: “The critical digital infrastructure sector continues to expand and to attract enviable levels of new investment. The ongoing build-out of new data centers and networks is largely being driven by cloud, hosted, and other as-a-service workloads…”
Overall, the critical infrastructure sector has adapted well to the pandemic. Fears of major outages, equipment shortages, huge budgetary overspends and a high loss of available staff have largely dissipated. These problems have been both more chronic and milder than feared — balanced by a strong uptick in demand as more people and services move online and more investment is made in resiliency and automation.
At the beginning of each calendar year, Uptime Institute highlights a short list of trends — or predictions — for the year ahead. The trends spanned growing demand for more resiliency and accountability; renewed investment in remote monitoring and automation; a belated investment wave in edge data center capacity; a significant increase in sustainability reporting requirements; and, finally, a wave of innovation. And for 2022? The Institute’s five predictions are unusual in that our confidence level in these predicted trends is not as high as usual, and some are also more contentious. However, the purpose of a prediction (or of highlighting a new trend) is not to be exact, but to provide some insights that may help to inform and guide the industry.
Sustained advances in chip technology over the past few decades, along with leaps in networking speeds, have helped the data center industry to grow. This has largely been through increased demand for IT services, but also because of vastly improved power efficiencies.
Moore’s law resumes — but not for all
In the past few years, however, the continuous improvements in chip processing speeds and efficiency have stalled as key suppliers, operating at the limits of physics, pushed their manufacturing capability to breaking point. Intel, the global leader in
server processors, with over 80% of the market, misjudged some crucial manufacturing technology choices — at a time when there was an insufficient alternative supply. This meant that while raw computing performance per server kept rising, the rate of improvement in IT energy efficiency slowed down — with potentially negative effects on global data center footprint, energy use and costs.
This is likely to change from 2022 onward, with an upswing in performance. However, there will not be a return to past dynamics. Some data centers will be able to exploit the new technology and make big gains in processing power and energy use — and some will not.
A major development is that Intel, whose dominance drives the industry, is set to find its mojo again — but only gradually. In the first months of 2022, the company will start making a new server processor generation, built on an enhanced version of its 10-nanometer technology (branded as Intel 7), to increase performance density. Intel’s 2022 server processors will bring improvements in efficiency too — but this will only be the first step in Intel’s technology comeback.
By the end of 2022, however, the market will have changed — in a way that some will be able to exploit far better than others. Regardless of how fast Intel recovers from its missteps, it will not regain its former grip on server processors, and data center operators will benefit (directly or indirectly) from a more diverse supply of chips. A resurgent Advanced Micro Devices (AMD), Intel’s long-standing rival, has already gained a recent foothold with processors that outperform those of its bigger rival’s products across a range of server applications.
These processors are delivering much higher performance more efficiently and, during 2022, AMD is expected to introduce increasingly efficient products. Other chip vendors, such as workload-acceleration specialist NVIDIA, as well as some hyperscale operators that have started building their own cloud server platforms — including Amazon Web Services (AWS) and Chinese cloud giant Alibaba — will be
migrating to new chip fabricating technologies with their own latest processor designs.
Underpinning these developments are TSMC (Taiwan Semiconductor Manufacturing Company) and Samsung Electronics (headquartered in South Korea), two foundries with cutting-edge manufacturing capabilities.
Industry consensus on sustainability looks fragile
Pressed by a sense of urgency among scientists and the wider public, and by governments and investors who must fulfil promises made at COP (Conference of the Parties) summits, major businesses are facing ever more stringent sustainability reporting requirements. Big
energy users, such as data centers, are in the firing line.
Much of the reporting requirements, and proposed methods of reducing carbon emissions, are going to be complicated and may appear contradictory and counterproductive. For this reason, we can safely predict that many managers will be bewildered and frustrated with the requirements to reduce and report greenhouse gas emissions and other environmental impacts in the years ahead. There are likely to be
disagreements between the various parties operating in (and / or overseeing) the data
center industry.
Most managers in the industry recognize the need for action, but trust in regulators is low. Asked if the data center sector needs greater regulation to improve its environmental sustainability, 63% of respondents to a recent Uptime Institute survey (Uptime Institute
Climate Change Survey 2021) said “yes.” But when asked how knowledgeable or informed regulators are in guiding and setting the rules for data center sustainability, respondents to this same survey gave regulators a low competency score.
To date, most of the commitments on climate change made by the industry have been voluntary. This has allowed a certain laxity in the definitions, targets and terminology used — and in the level of scrutiny applied. But these are all set to be tested: reporting
requirements will increasingly become mandatory, either by law or because of commercial pressures. Failure to publish data or meet targets will carry penalties or have other negative consequences.
Data center operators ponder the nuclear option
Data center owners and operators worldwide have long been under pressure to use sustainable power — with little to no carbon emissions. Despite a decade of effort, few data center operators are yet to lay claim to having even one data center that uses entirely carbon-free energy.
This is the struggle that has been faced by all businesses worldwide. Both renewable energy generation and emission reduction pledges at the 2021 COP26 summit in Glasgow, Scotland, fell well short of what is required to keep global warming below two
degrees Celsius. In 2021, power generation from coal jumped to meet a rising demand for energy — with associated carbon emissions increasing by 3.5%, the International Energy Agency projected.
As major businesses feel a growing sense of urgency to dramatically cut carbon emissions, opinions are starting to shift in favor of nuclear power, which is not classed as clean, but is a near-zero carbon energy source. The digital infrastructure industry, a major global
consumer of energy, has a role to play in rehabilitating nuclear, and in marrying it to intermittent renewable energy to provide firm zero-emission power.
There is considerable reluctance to use, or endorse, the use of nuclear power, largely stemming from a fear of meltdowns and concerns about nuclear waste. These worries are likely to be overcome by the need for dispatchable, low-carbon power generation
that does not depend on local weather. From 2022, we expect some major data center operators, and industry influencers and leaders, to support nuclear power more actively and openly — even pressuring governments and utilities to invest in this option.
The Uptime Institute sees three areas where the data center industry will (or may) have an active role in supporting a significant shift toward nuclear power. Two of these areas are important and immediate; the third, involving the development of new nuclear technologies, is more
speculative.
Concerns over cloud concentration risk grow
The demand for cloud computing was already experiencing rapid growth, but this growth accelerated during the pandemic as more businesses and consumers became increasingly dependent on an ever-expanding online services. While a public cloud provides a flexible, stable and distributed IT environment, there are growing concerns around its use.
These issues center on resiliency, an area almost all public cloud providers can fairly claim to excel. Organizations with mission-critical workloads build in redundancy at every level of the IT and have processes to work around failures when they do occur. Following
some recent high-profile cloud failures, and with regulators asking more questions, there is growing concern that using a big cloud provider is a single point of failure, not just technically but also from a business-risk perspective.
Many organizations and regulators are concerned with the lack of transparency of cloud providers, and the lack of control that important clients have — some of which are part of the national critical infrastructure. Concentration risk, where key services are dependent on one or a few key suppliers, is a particular concern.
Supply problems favor standardization and scale
The COVID-19 pandemic has stressed supply chains globally, resulting in higher prices and increased delivery times for many components. Shortages and delays are likely to persist into 2022 and beyond as demand for IT, and for new data center capacity,
continues to increase.
To keep pace, data center owners and their suppliers are taking steps to mitigate supply chain disruption — but the playing field is far from level. The biggest operators, and especially those with a multinational operation, are using their buying-power, relationships and scale to secure supplies, while smaller ones may struggle. This will further widen the cost advantage that bigger data center builders and operators already enjoy.
Since the start of the pandemic, delivery times for certain critical data center equipment, including intelligent breakers, cooling components and uninterruptible power supply (UPS) systems, have increased from one to six months in many regions. Delays are often
due to a shortage of parts, such as semiconductors, compressors or fans. Supply chain disturbances have led to higher prices for silicon and other raw materials (including copper), as well as for transportation of components by both sea and land.
Despite these issues, there remains an overall expectation that the total cost and time to deliver new data center capacity (per MW of IT) will continue to decrease, as it has done for many years (see the Uptime Institute report Best-in-class data center provisioning).
To meet expectations for faster, more cost-efficient builds, data center operators are collaborating with their key suppliers more closely, including with general contractors and system / component vendors. The biggest cloud and internet giants have been
working with suppliers to value engineer products and components for many years, but this collaboration now includes component availability. This might mean, for example, densification (raising the IT load per rack to reduce the facility footprint and cost) and increasing the use of off-site system integration, such as making use of electrical power skids, containerized cooling systems and other prefabricated components.
