Despite their role as enablers of technological progress, data center operators have been slow to take advantage of developments in software, connectivity and sensor technologies that can help optimize and automate the running of critical infrastructure.
Most data center owners and operators currently use building management systems (BMS) and / or data center infrastructure management (DCIM) software as their primary interfaces for facility operations. These tools have important roles to play in data center operations but have limited analytics and automation capabilities, and they often do little to improve facility efficiency.
To handle the increasing complexity and scale of modern data centers — and to optimize for efficiency — operations teams need new software tools that can generate value from the data produced by the facilities equipment.
For more than a decade, DCIM software has been positioned as the main platform for organizing and presenting actionable data — but its role has been largely passive. Now, a new school of thought on data center management software is emerging, proposed by data scientists and statisticians. From their point of view, a data center is not a collection of physical components but a complex system of data patterns.
Seen in this way, every device has a known optimal state, and the overall system can be balanced so that as many devices as possible are working in this state. Any deviation would then indicate a fault with equipment, sensors or data.
This approach has been used to identify defective chiller valves, imbalanced computer room air conditioning (CRAC) fans, inefficient uninterruptible power supply (UPS) units and opportunities to lower floor pressure — all discovered in data, without anyone having to visit the facility. The overall system — the entire data center — can also be modelled in this way, so that it can be continuously optimized.
All about the data
Data centers are full of sensors. They can be found inside CRAC and computer room air handler units, chillers, coolant distribution units, UPS systems, power distribution units, generators and switchgear, and many operators install optional temperature, humidity and pressure sensors in their data halls.
These sensors can serve as a source of valuable operational insight — yet the data they produce is rarely analyzed. In most cases, applications of this information are limited to real-time monitoring and basic forecasting.
In recent years, major mechanical and electrical equipment suppliers have added network connectivity to their products as they look to harvest sensor data to understand where and how their equipment is used. This has benefits for both sides: customers can monitor their facilities from any device, anywhere, while suppliers have access to information that can be used in quality control, condition-based or predictive maintenance, and new product design.
The greatest benefits of this trend are yet to be harnessed. Aggregated sensor data can be used to train artificial intelligence (AI) models with a view to automating an increasing number of data center tasks. Data center owners and operators do not have to rely on equipment vendors to deliver this kind of innovation. They can tap into the same equipment data, which can be accessed using industry-standard protocols like SNMP or Modbus.
When combining sensor data with an emerging category of data center optimization tools — many of which rely on machine learning — data center operators can improve their infrastructure efficiency, achieve higher degrees of automation and lower the risk of human error.
The past few years have also spawned new platforms that simplify data manipulation and analysis. These enable larger organizations to develop their own applications that leverage equipment data — including their own machine learning models.
The new wave
Dynamic cooling optimization is the best-understood example of this new, data-centric approach to facilities management. These applications rely on sensor data and machine learning to determine and continually “learn” the relationships between variables such as rack temperatures, cooling equipment settings and overall cooling capacity. The software can then tweak the cooling equipment performance based on minute changes in temperature, enabling the facility to respond to the needs of the IT in near real-time.
Many companies working in this field have close ties to the research community. AI-powered cooling optimization vendor TycheTools was founded by a team from the Technical University of Madrid. The Dutch startup Coolgradient was co-founded by a data scientist and collaborates with several universities. US-based Phaidra has brought together some of the talent that previously published and commercialized cutting-edge research as part of Google’s DeepMind.
Some of the benefits offered by data-centric management software include:
Improved facility efficiency: through the automated configuration of power and cooling equipment as well as the identification of inefficient or faulty hardware.
Better maintenance: by enabling predictive or condition-based maintenance strategies that consider the state of individual hardware components.
Discovery of stranded capacity: through the thorough analysis of all data center metrics, not just high-level indicators.
Elimination of human error: through either a higher degree of automation or automatically generated recommendations for human employees.
Improvements in skill management: by analyzing the skills of the most experienced staff and codifying them in software.
Not all machine learning models require extensive compute resources, rich datasets and long training times. In fact, many of the models used in data centers today are small and relatively simple. Both training and inference can run on general-purpose servers, and it is not always necessary to aggregate data from multiple sites — a model trained locally on a single facility’s data will often be sufficient to deliver the expected results.
New tools bring new challenges
The adoption of data-centric tools for infrastructure management will require owners and operators to recognize the importance of data quality. They will not be able to trust the output of machine learning models if they cannot trust their data — and that means additional work on standardizing and cleaning their operational data stores.
In some cases, data center operators will have to hire analysts and data scientists to work alongside the facilities and IT teams.
Data harvesting at scale will invariably require more networking inside the data center — some of it wireless — and this presents a potentially wider attack surface for cybercriminals. As such, cybersecurity will be an important consideration for any operational AI deployment and a key risk that will need to be continuously managed.
Evolution is inevitable
Uptime Institute has long argued that data center management tools need to evolve toward greater autonomy. The data center maturity model (see Table 1) was first proposed in 2019 and applied specifically to DCIM. Four years later, there is finally the beginning of a shift toward Level 4 and Level 5 autonomy, albeit with a caveat — DCIM software alone will likely never evolve these capabilities. Instead, it will need to be combined with a new generation of data-centric tools.
Table 1. The Data Center Management Maturity Model
Not every organization will, or should, take advantage of Level 4 and Level 5 functionality. These tools will provide an advantage to the operators of modern facilities that have exhausted the list of traditional efficiency measures, such as those achieving PUE values of less than 1.3.
For the rest, being an early adopter will not justify the expense. There are cheaper and easier ways to improve facility efficiency that do not require extensive data standardization efforts or additional skills in data science.
At present, AI and analytics innovation in data center management appears to be driven by startups rather than established software vendors. Few BMS and DCIM developers have integrated machine learning into their core products, and while some companies have features in development, these will take time to reach the market — if they ever leave the lab.
Uptime Intelligence is tracking six early-stage or private companies that use facilities data to create machine learning models and already have products or services on the market. It is likely more will emerge in the coming months and years.
These businesses are creating a new category of software and services that will require new types of interactions with all the moving parts inside the data center, as well as new commercial strategies and new methods of measuring the return on investment. Not all of them will be successful.
The speed of mainstream adoption will depend on how easy these tools will be to implement. Eventually, the industry will arrive at a specific set of processes and policies that focus on benefitting from equipment data.
The Uptime Intelligence View
The design and capabilities of facilities equipment have changed considerably over the past 10 years, and traditional data center management tools have not kept up. A new generation of software from less-established vendors now offers an opportunity to shift the focus from physical infrastructure to data. This introduces new risks — but the benefits are too great to ignore.
https://journal.uptimeinstitute.com/wp-content/uploads/2024/08/Data-center-management-software-is-evolving-at-last-featured.jpg5401030Max Smolaks, Research Analyst, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngMax Smolaks, Research Analyst, [email protected]2024-08-21 15:00:002024-08-19 15:24:23Data center management software is evolving — at last
In the past year, Uptime Intelligence has been asked more questions about generative AI and its impact on the data center sector than any other topic. The questions come from enterprise and colocation operators, suppliers of a wide variety of equipment and services, regulators and the media.
Most of the questions concern power consumption. The compute clusters — necessary for the efficient creation and use of generative AI models — are enormously power-hungry, creating a surge in projected demand for capacity and, for operators, challenges in power distribution and cooling in the data center.
The questions about power typically fall into one of three groups. The first is centered around, “What does generative AI mean for density, power distribution and cooling in the data center?”
Uptime Intelligence’s view is that the claims of density distress are exaggerated. While the AI systems for training generative AI models, equipped with Nvidia accelerators, are much denser than usual, they are not extreme and can be managed by spreading them out into more cabinets. Further Uptime Intelligence reports will cover this aspect since it is not the focus here.
Also, generative AI has an indirect effect on most data center operators in the form of added pressures on the supply chain. With lead times on some key equipment, such as engine generators, transformers and power distribution units, already abnormally long, unforeseen demand for even more capacity by generative AI will certainly not help. The issues of density and forms of indirect impact were initially addressed in the Uptime Intelligence report Hunger for AI will have limited impact on most operators.
The second set of questions relates to the availability of power in a given region or grid — and especially of low-carbon energy. This is also a critical, practical issue that operators and utilities are trying to understand. The largest clusters for training large generative AI models comprise many hundreds of server nodes and draw several megawatts of power at full load. However, the issues are largely localized, and they are also not the focus here.
Generative AI and global power
Instead, the focus of this report is the third typical question, “How much global power will AI use or require?” While in immediate practical terms this is not a key concern for most data center operators, the headline numbers will shape media coverage and public opinion, which ultimately will drive regulatory action.
However, some of the numbers on AI power circulating in the press and at conferences, cited by key influencers and other parties, are extremely high. If accurate, these figures suggest major infrastructural and regulatory challenges ahead — however, any unduly high forecasts may prompt regulators to overreact.
Among the forecasts at the higher end is from Schneider Electric’s respected research team, which estimated AI power demand at 4 gigawatts (GW), equivalent to 35 terawatt-hours (TWh) if annualized, in 2023, rising to around 15 GW (131 TWh if annualized) in 2028 (see Schneider White paper 110: The AI Disruption: Challenges and Guidance for Data Center Design). Most likely, these figures include all AI workloads and not only new generative models.
And Alex de Vries, of digital trends platform Digiconomist and whose calculations of Bitcoin energy use have been influential, has estimated AI workload use at 85 TWh to 134 TWh by 2027. These figures suggest AI could add 30% to 50% or more to global data center power demand over the next few years (see below).
There are two reasons why Uptime Intelligence considers these scenarios overly bullish. First, estimates on power for all AI workloads are problematic for both taxonomy (what is AI) and the improbability of tracking it meaningfully. Also, most forms of AI are already accounted for in capacity planning, despite generative AI being unexpected. Second, projections that span beyond a 12- to 18-month horizon carry high uncertainty.
Some of the numbers cited above imply an approximately thousand-fold increase in AI compute capacity in 3 to 4 years from the first quarter of 2024, when accounting for hardware and software technology evolution. That is not only unprecedented but also has weak business fundamentals when considering the hundreds of billions of dollars it would take to build all that AI infrastructure.
Uptime Intelligence takes a more conservative view with its estimates — but these estimates still indicate rapidly escalating energy use by new large generative AI models.
To reach our numbers, we have estimated the current shipments and installed base of Nvidia-based systems through to the first quarter of 2025 and the likely power consumption associated with their use. Systems based on Nvidia’s data center accelerators, derived from GPUs, dominate the generative AI model accelerator market and will continue to do so until at least mid-2025 due to an entrenched advantage in the software toolchain.
We have considered a range of technical and market factors to calculate the power requirements of generative AI infrastructure: workload profiles (workload activity, utilization and load concurrency in a cluster), the shifting balance between training and inferencing, and the average PUE of the data center housing the generative AI systems.
This data supports our earlier stated view that the initial impact of generative AI is limited, beyond a few dozen large sites. For the first quarter of 2024, we estimate the annualized power use by Nvidia systems installed to be around 5.8 TWh. This figure, however, will rise rapidly if Nvidia meets its forecast sales and shipment targets. By the first quarter of 2025, the generative AI infrastructure in place could account for 21.9 TWh of annual energy consumption.
We expect these numbers to shift as new information emerges, but they are indicative. To put these numbers into perspective, the total global data center energy use has been variously estimated at between 200 TWh and 450 TWh per year in the periods from 2020 to 2022. (The methodologies and terms of various studies vary widely and suggest that data centers use between 1% and 4% of all consumed electricity.) By taking a middle figure of 300 TWh for the annual global data center power consumption, Uptime Intelligence puts generative AI annualized energy at around 2.3% of the total grid power consumption by data centers in the first quarter of 2024. However, this could reach 7.3% by the first quarter of 2025.
Outlook
These numbers indicate that generative AI’s power use is not, currently, disruptively impactful, given the data center sector’s explosive growth in recent years. Generative AI’s share of power consumption relative to its footprint is outsized given its likely very high utilization. A share of data center capacity by power will fall in the low-to-mid single digits even by the end of 2025.
However, it is nevertheless a dramatic increase and suggests that generative AI could account for a much larger percentage of overall data center power use in the years ahead. Inevitably, the surge needs to slow down, also in great part because newer AI systems built around vastly more efficient accelerators will displace the install base en masse rather than adding net new infrastructure.
While Uptime Intelligence believes some of the estimates of generative AI power use (and data center power use) to be too high, the sharp uptick — and the concentration of demand in certain regions — will still be high enough to attract and stimulate regulatory attention. Uptime Intelligence will continue to analyze the development of AI infrastructure and its impact on the data center industry.
https://journal.uptimeinstitute.com/wp-content/uploads/2024/07/Generative-AI-and-global-power-consumption-high-but-not-that-high-featured.jpg5401030Andy Lawrence, Executive Director of Research, Uptime Institute, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngAndy Lawrence, Executive Director of Research, Uptime Institute, [email protected]2024-07-31 12:00:002024-07-31 10:03:53Generative AI and global power consumption: high, but not that high
Uptime Intelligence regularly addresses IT infrastructure efficiency, particularly servers, in our reports on data center energy performance and sustainability. Without active contribution from IT operations, facility operations alone will not be able to meet future energy and sustainability demands on data center infrastructure. Purchases of renewable energy and renewable energy certificates will become increasingly — and, in many locations, prohibitively — expensive as demand outstrips supply, making the energy wasted by IT even more costly.
The power efficiency of a server fleet, that is, how much work servers perform for the energy they use, is influenced by multiple factors. Hardware features receive the most attention from IT buyers: the server’s technology generation, the configuration of the system and the selection of power supply or fan settings. The single most significant factor that affects server efficiency, however, is the level at which the servers are typically utilized; a seemingly obvious consideration — and enough for regulators to include it as a reporting requirement in the EU’s new Energy Efficiency Directive (see EED comes into force, creating an enormous task for the industry). Even so, the process of sourcing the correct utilization data for the purposes of power efficiency calculations (as opposed to capacity planning) remains arguably misunderstood (see Tools to watch and improve power use by IT are underused).
The primacy of server utilization in data center efficiency has increased in recent years. The latest server platforms are only able to deliver major gains in energy performance when put to heavy-duty work — either by carrying a larger software payload through workload consolidation, or by running scalable, large applications. If these conditions are not met, running lighter or bursty workloads on today’s servers (regardless of whether based on Intel or AMD chips) will deliver only a marginal, if any, improvement in the power efficiency compared with many of the supposedly outdated servers that are five to seven years old (see Server efficiency increases again — but so do the caveats).
Cycles of a processor’s sleep
This leads into the key discussion point of this report: the importance of taking advantage of dynamic energy saving features. Settings for power and performance management of servers are often an overlooked — and underused — lever in improving power efficiency. Server power management techniques affect power use and overall system efficiency significantly. This effect is even more pronounced for systems that are only lightly loaded or spend much of their time doing little work: for example, servers that run enterprise applications.
The reduction in server power demand resulting from power management can be substantial. In July 2023 Uptime Intelligence published a report discussing data (although sparse) that indicates 10% to 20% reductions in energy use from enabling certain power-saving modes in modern servers, with only a marginal performance penalty when running a Java-based business logic (see The strong case for power management). Energy efficiency gains will depend on the type of processor and hardware configuration, but we consider the results indicative for most servers. Despite this, our research indicates that many, if not most, IT operations do not use power management features.
So, what are these power management settings? Server power management settings are governed by the firmware statically (what modes are enabled upon system start up) and dynamically by the operating system or hypervisor once running through the Advanced Configuration and Power Interface.
There are many components in a server that may have power management features, enabling them to run slower or power off. Operating systems also have their own software mechanisms, such as suspending their operation and saving the machine state to central memory or the storage system.
But in servers, which tend to be always powered on, it is the processors’ power management modes that dictate most of the energy gains. Modern processors have sophisticated power management features for idling, that is, when the processor does not execute code. These are represented by various levels of C-states (the C stands for CPU) denoted by numbers, such as C1 and C2 (with C0 being the fully active state).
The number of these states has expanded over time as chip architects introduce new, more advanced power-saving features to help processors reduce their energy use when doing no work. The chief benefit of these techniques is to minimize leakage currents that would otherwise increasingly permeate modern processor silicon.
The higher the C-state number, the more of its circuitry the CPU sends to various states of sleep. In summary:
C0: processor active.
C1/C1E: processor core halts, not performing work, but is ready to immediately resume operation with negligible performance penalty, optionally reducing its voltage and frequency to save power.
C3: processor clock distribution is switched off and core caches are emptied.
C4: enhancement to C3 that extends the parts covered.
C6: essentially powers down entire cores after saving the state to resume from later.
C7 and higher: shared resources between cores may be powered downs, or even the entire processor package.
Skipped numbers, such as C2 and C5, are incremental, transitionary processor power states between main states. Not all these C-states are available on all processor architectures.
A good sleep in a millisecond
The levels of C-states and understanding them matters because they largely define the cost in performance and the benefit in power. The measured results of 10% to 20% reduction in energy use when enabling certain power management features, as discussed earlier, have allowed the server processor (an AMD model) to enter processor power states up to C6. These sleep states save power even when the server, on a human level of perception, is processing database transactions and responding to queries.
This is because processors operate on a timescale measured in nanoseconds, while software-level requests between commands can take milliseconds even on a busy machine. This is a factor of one million difference: milliseconds between work assignments represent millions of processor cycles waiting. For modern server processors, some of the many cores may often have no work to do for a second or more, which is an eternity on the processor’s time scale. On a human scale, a comparable time would be several years of inactivity.
However, there is a cost associated with the processor cores going to sleep. Entering ever deeper sleep states across processor cores or entire chips can take thousands of cycles, and as many as tens of thousands of cycles to wake up and reinstate operation. This added latency to respond to wake-up requests is what shows up as a loss of performance in measurements. In the reference measurement running the Java-based business logic, this is in the 5% to 6% range — arguably a small price to pay.
Workloads will vary greatly in the size of the performance penalty introduced by this added latency. Crucially, they will differ even more in how costly the lost application performance is for the business — high-frequency trading or processing of high volumes of mission-critical online transactions are areas where any loss of performance is unacceptable. Another area may include storage servers with heavy demand for handling random read-write operations at low latency. But a vast array of applications will not see material change to the quality of service.
Using server power management is not a binary decision either. IT buyers can also calibrate the depth of sleep they enable for the server processor (and other components) to enter. Limiting it to C3 or C1E may deliver better trade-offs. Many servers, however, are not running performance-critical applications and spend most of their time doing no work — even if it seems that they are often, by human standards, called upon. For servers that are often idle, the energy saved can be in the 20% to 40% range, which can amount to tens of watts for every lightly loaded or idle server.
Optimizing server energy performance does not stop with C-states. Performance-governing features (setting performance levels when the processor is actively working), known as P-states, offer another set of possibilities to find better trade-offs between power and performance. Rather than minimizing waste when the processor idles, P-states direct how much power should be expended on getting the work done. Future reports will introduce P-states for a more complete view of server power and performance management for IT infrastructure operators that are looking for further options in meeting their efficiency and sustainability objectives.
The Uptime Intelligence View
A server processor’s power management is a seemingly minute function buried under layers of technical details of an infrastructure. Still, its role in the overall energy performance of a data center infrastructure will be outsized for many organizations. In the near term, blanket policies (or simply IT administrator habits) of keeping server power management features switched off will inevitably be challenged by internal stakeholders in pursuit of cost efficiencies and better sustainability credentials; or, possibly in the longer term, by regulators catching on to technicalities and industry practices. Technical organizations at enterprises and IT service providers will want to map out server power management opportunities ahead of time.
https://journal.uptimeinstitute.com/wp-content/uploads/2024/07/Understanding-how-server-power-management-works-featured.jpg5401030Daniel Bizo, Research Director, Uptime Institute Intelligence, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngDaniel Bizo, Research Director, Uptime Institute Intelligence, [email protected]2024-07-10 15:00:002024-07-09 10:40:07Understanding how server power management works
The European Green Deal, a set of policy initiatives approved in 2020, aims for a sustainable and competitive economy with net-zero greenhouse gas emissions by 2050. Along with the legislation driving this transition, such as the Energy Efficiency Directive recast (see EED comes into force, creating an enormous task for the industry), the strategy will require much higher rates of electrification in transport and industries to displace the use of fossil fuels. This aim is combined with the policy objective of adding large amounts of renewable power generation capacity.
However, this direction is leading toward a new environmental challenge: increased sales of electric vehicles (EVs) will create future end-of-life battery problems, drawing attention to the unresolved issue of reusing and recycling large battery banks. Many industrial stationary applications are also seeing a strong take-up of EV-type (mostly lithium-ion) batteries for a variety of reasons. These include the displacement of valve-regulated lead-acid (VRLA) batteries, which are highly recycled, new energy storage installations for grid demand-response schemes and the elimination of standby engine generators.
Until now, this area has been governed by the 2006 Battery Directive (2006/66/EC). However, this directive is being replaced by Regulation 2023/1542 — the main objective of which is to create an updated set of rules to ensure high sustainability standards regardless of the battery chemistry. As an EU Regulation, it applies automatically to all member states without implementation in national laws.
The new regulations address aspects such as carbon footprint, recycled content, safety, labeling and end-of-life management. Industrial customers, including data center suppliers and their customers, will need to start reporting by mid-2025 — although this date may yet change. Although these rules only apply to members of the EU, its standards and laws are often replicated in other countries.
The 2006 Battery Directive was put in place to mitigate the environmental impact of battery production and disposal of waste batteries. It introduced recycling and treatment targets, restrictions on some hazardous substances, labeling requirements (for hazardous substances and instructions for proper disposal), reporting obligations and extended producer responsibility.
Lead-acid versus Li-ion recycling: the facts
The most common batteries used in uninterruptible power supply (UPS) systems are VRLA and Li-ion batteries (of which several sub-types exist). Lead-acid batteries have the highest collection and recycling rates. In the EU, the recycling rate for automotive starter batteries is 99% and more than 90% of the lead is recovered. The figure for stationary applications, which includes data centers, is likely to be similarly high due to the stringent regulations and economic incentives to recycle.
In 2021, all EU member states met the target recycling rate of 65% by weight for lead-acid batteries (both automotive and non-automotive).
The recycling process of lead-acid batteries consists of draining the electrolyte, opening the casing and separating the materials. The lead plates are then smelted to obtain molten lead, which is purified and refined before being cast into ingots for reuse. The plastic components are also recycled. Lead-acid batteries have few components and contain approximately 70% lead, which means that it is an efficient process. It is also profitable because recycled lead can be used in the manufacture of new batteries and is recyclable at relatively low temperatures, which require less energy.
For Li-ion batteries, the view is more complicated. Currently, it is cheaper to mine the metals to make new Li-ion cells (regardless of Li-ion chemistry) than it is to recycle used batteries. As a result, the much higher environmental footprint of the supply chain to produce Li-ion cells has so far been an externality and not priced into the cost of making Li-ion batteries. This factors heavily into its current market advantage in price-performance compared to other, environmentally less damaging cell chemistries.
The two most common processes for recycling spent Li-ion batteries are pyrometallurgy and hydrometallurgy. During both processes, the batteries are discharged and dismantled. They are then either smelted at high temperatures or leached to recover high-value cathode materials, such as cobalt, nickel and copper. During pyrometallurgical smelting, the lithium is lost in the furnace, making the process for lithium not sustainable. Future recycling processes for Li-ion cells are in development and will likely involve a combination of heat treatment and leaching in various concentrations of acid to maximize recovery of valuable metals.
These processes are energy-intensive and therefore expensive to operate, as well as producing gaseous pollutants and industrial waste. According to some battery vendors, the costs involved with the transport and recycling of Li-ion batteries, which is considered a hazardous waste, are substantial.
This is a major reason behind the current expectation that Li-ion battery packs will find secondary and even tertiary uses rather than being recycled for their raw materials. Compared with lead-acid batteries, this is possible due to Li-ion cells’ higher endurance and shelf life. The new battery regulation actively considers this scenario with the introduction of battery passports.
What are the new rules?
The phased implementation of the rules (Regulation 2023/1542) begins in July 2024 and regulates the carbon footprint, recycled content of new batteries, labeling and the introduction of an online battery information system. The new battery regulation controls all battery chemistries, with rules varying by battery category, for example, EV, industrial and portable. Recycling targets differ between chemistries, with specific targets for the recovery of cobalt, lead, lithium and nickel. Figure 1 sets out the major milestones of the regulatory rollout for EU member states, as prescribed in the regulation.
Figure 1. Timeline of the EU’s battery regulation
New batteries put to market will be subject to mandatory minimum levels of recycled content requirements. From 2030, batteries will need to contain a minimum recycled content of 12% for cobalt, 4% for lithium, 4% for nickel and 85% for lead. By 2035, these thresholds will increase to 20% cobalt, 10% lithium, 12% nickel and 85% lead.
Alongside these requirements, there are also recycling efficiency and material recovery targets for end-of-life batteries. By the end of 2030, used batteries will have a recycling target by weight of 80% for lead-acid and 70% for Li-ion. The material recovery target is 95% for cobalt, copper, lead and nickel and 70% for lithium.
What does this mean for data center operators?
The new rules will mostly affect battery producers; however, some responsibilities will fall on end users, such as data center operators. These include performing due diligence to verify vendor (product carbon footprint) and recycler claims (environmental credentials of the recycling process), as well as maintaining up-to-date battery passports during the use of the products.
From August 18, 2025, battery suppliers and data center operators (with some exceptions), will have a legal requirement to adopt a battery due diligence policy covering the social and environmental risks. This policy will need to include annual reports on the risks in the supply chain and the steps taken to manage them, transparency on sourcing raw and recycled materials, and structuring an internal management system to support it. The due diligence policy should be verified by a third party and communicated to suppliers and the public via annual reports.
There are exceptions for suppliers and buyers of batteries with an annual net turnover of less than €40 million ($43 million) and those using batteries that have undergone preparation for reuse, repurposing, or remanufacturing before being placed on the market.
The introduction of battery passports in 2026 will affect both suppliers and end users, including data center operators. Battery producers will be required to report on properties such as the chemistry, carbon footprint and recycled content, while it will be the end user’s responsibility to update the battery passport on the state of its health throughout its life. This will likely include maintenance entries but also any substantial events or accidents, such as deviations from standard operating environmental conditions, overcharging or deep discharging, and other impacts.
Battery passports will be key to enabling a second-hand market for batteries that are appropriate for reuse. Alongside the labeling of batteries, this may lead to improvements in Li-ion battery recycling by simplifying the separation process by sorting batteries by chemistry before the start of the recycling process. However, the technical implementation of the battery passport has not been stipulated in the new regulation and will be left to future cooperation between EU member states.
The regulation states that producers shall cover the necessary costs incurred by the collection and recycling of waste batteries. Lead-acid batteries have an inherent economic value at the end of their useful lives, which guarantees incentives for both buyers and sellers to promote recycling. Li-ion batteries, however, currently incur substantial costs to pay for recycling. According to battery makers, this can amount to as much as the price of new Li-ion batteries. Another potentially overlooked detail is the cost of transporting large amounts of Li-ion batteries that are classified as hazardous and pose a fire risk.
It is possible (or even likely) that end users may see a marked increase in the price of Li-ion batteries — the selling price will need to cover the recycling costs as a result of this extended producer responsibility. Data center operators, with hundreds of batteries in their UPS systems, may have to set up databases and processes to ensure they are reliably tracked.
The regulation stipulates that the ultimate responsibility for managing end-of-life batteries will fall on suppliers — in practical terms: taking batteries from end-users for no charge and ensuring the batteries are reused or recycled. However, end users should also exercise due diligence, even when they are not legally required to at present. Data center owners and operators should be aware that battery manufacturers may overstate their recycling capabilities. This is particularly relevant given the long lifespan of industrial Li-ion batteries, which can remain in use for 10 to 15 years or longer. Companies placing such batteries on the market may not have adequate recycling plans in place.
End users should also consider what technology is used in the recycling facilities, and even verify the existence of these facilities by checking their reported location using satellite imagery and public records. Inevitably, some battery vendors will go out of business, leaving end users with a potentially costly liability.
Additionally, when considering Scope 3 emissions reporting (in accordance with the Greenhouse Gas Protocol), the new recycling obligations add complexity. End users will need to consider carbon emissions from transportation and recycling processes. While some may argue that recycling offsets carbon emissions compared with manufacturing new batteries from raw materials, this claim may not withstand closer scrutiny.
Where could these rules go in the future?
The 2023 battery regulation provides a timeline of implementation, but the rules will need further clarification when they start to come into effect. Potential secondary legislation could involve standardizing calculations for carbon footprint, recycling efficiency and material recovery. More clarification may follow for the implementation and functioning of the electronic exchange system as well as specifications for the content and accessibility of battery passports.
Additionally, EU member states may choose to enact further requirements (as long as they are not in conflict with EU law) to stimulate investments in research and development related to carbon footprint reduction and sustainability in battery production and recycling.
The Uptime Intelligence View
The new EU Battery Regulation is primarily a response to the mass-market adoption of EVs, however, it also covers industrial stationary applications, such as mission-critical power systems.
The EU’s objective is to ensure that huge quantities of new batteries will not simply end up as hazardous waste at the end of their lives but will either find new uses or be recycled to make new battery cells. It will also level the playing field with lead-acid batteries and other, more readily recyclable chemistries.
Rosa Lawrence, Research Associate, Uptime Institute, [email protected]
https://journal.uptimeinstitute.com/wp-content/uploads/2024/06/EU-battery-regulations-what-do-the-new-rules-mean-featured.jpg5401030Rosa Lawrence, Research Associate, Uptime Institute, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngRosa Lawrence, Research Associate, Uptime Institute, [email protected]2024-06-20 15:00:002024-06-20 09:40:06EU battery regulations: what do the new rules mean?
The colocation and public cloud sectors of the digital infrastructure industry continue to make headlines, with many organizations planning large-scale capacity expansion to meet rising demand. However, there is also a less public expansion underway — enterprises operators, for the third successive year, say they are going to invest in more data center capacity in 2024.
Results from the Uptime Institute Capacity Trends Survey 2023 reveal that 64% of enterprise operators are growing their data center capacity — a six percentage-point uptick from two years earlier in 2021 (see Figure 1). Notably, one in five organizations in this group say they are expanding by more than 20% annually. This scale of expansion is difficult to implement without significant investment.
Figure 1. Enterprise data center capacity shows strong growth
Some suppliers — and some operators — may be surprised by this rate of growth since it follows a decade-long period in which enterprise data centers were often dismissed as expensive, inflexible and outdated by many executives. However, Uptime’s data, which has been consistent over the past three years and is backed by many conversations, suggests that enterprise investment is strong.
What is driving this growth in enterprise data center capacity? Partly, it is simply a demand for more digital services. However, companies also say that they are investing to enhance the resiliency of their data centers (45% of survey respondents) and to support a hybrid cloud strategy (37%). Moving to cloud architectures often requires an increase in data center capacity, especially if an organization is developing distributed resiliency architectures.
Cost may also be a factor that is driving enterprise investment. According to separate Uptime research, of those enterprise operators that compared the cost of provisioning workloads on-premises versus off-premises, most report that corporate data centers are less expensive than using colocation (56%, n=154) or public cloud (51%, n=151). This is particularly true for enterprises that have already made significant investments to expand capacity.
One looming problem for enterprises — and indeed for colocation companies — is the widely forecast increase in rack power density in the coming years. To accommodate this, new investments in cooling and power distribution will be required.
Four-fifths (82%) of enterprises say that they are expecting more demand for higher power densities in the next two to three years, but more than one-third (36%) say that they cannot accommodate this demand with their existing infrastructure (see Figure 2). As a result, many workloads with higher density demands will, as Uptime expects, be outsourced to third parties that have the requisite power and cooling infrastructure.
Figure 2. Many need new investment to meet expected power densities
In spite of the increased enterprise sector spending, the trend towards greater outsourcing to colocation and cloud companies is expected to remain strong (see The majority of enterprise IT is now off-premises). For example, colocation companies report more growth than their enterprise counterparts by 15 percentage points (79%, n=130), with twice as many reporting annual growth rates of more than 20% (40%, n=130).
Taken together, Uptime’s survey data shows that chief information officers are investing in cloud, hosting, colocation and enterprise data centers. While more workloads may be outsourced, the enterprise data center will most likely continue to grow and evolve. Large companies with complex mission-critical workloads, especially those that are heavily regulated, will most likely maintain on-premises sites.
However, as colocation and public cloud providers expand the depth of their services in response to the industry’s staffing, regulatory and supply chain challenges, enterprises will increasingly integrate these resources over the next decade.
The Uptime Intelligence View
Enterprise data centers have been characterized as being in decline over the past five years, especially in the context of the significant, double-digit annual growth of large colocation and public cloud organizations. But Uptime survey data has consistently shown investment in the sector. Although owning and operating data centers may not feature in the strategies employed by many large and especially newer organizations, enterprise facilities will likely remain essential to businesses beyond the medium term.
https://journal.uptimeinstitute.com/wp-content/uploads/2024/06/Colocation-and-public-cloud-growth-masks-enterprise-expansion-featured.jpg5401030Douglas Donnellan, Senior Research Associate, Uptime Institute, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngDouglas Donnellan, Senior Research Associate, Uptime Institute, [email protected]2024-06-05 15:00:002024-06-03 17:19:21Colocation and public cloud growth masks enterprise expansion
Human error has been — and remains to be — a major cause of outages in data centers. Uptime Intelligence’s research shows that about four in 10 operators have had a major outage in the past three years in which human error played a role (Annual outage analysis 2023). Half of these respondents said errors were made because staff failed to follow the correct procedures.
Thorough training, regular practice in equipment testing and work experience all help to reduce these errors — particularly in an emergency when a prompt reaction is crucial. An often underappreciated factor is the importance of mental performance and the effects of fatigue.
The relationship between shift length, fatigue and human error is well documented, but less clear is how the data center industry can define shifts that help minimize human error. The recommended best practices for other industries do not always translate into the data center world, where 24/7 service availability is the standard. Additionally, data center owners and operators wanting to optimize shift length to limit fatigue need to navigate employee preferences and region-specific constraints.
What the research says
Studies indicate there is a tipping point after which the performance of most staff deteriorates. Researchers at the Chinese University of Hong Kong Department of Systems Engineering and Engineering Management analyzed 241 papers on the relationship between shift length and occupational health and found that individuals working more than 10-hour shifts are significantly more likely to experience fatigue. A similar review from the Finnish Institute of Occupational Health shows the risk of workplace injury due to fatigue-related accidents across a range of industries is 15% higher in 10-hour shifts than 8-hour shifts, and jumps to 38% higher at 12 hours.
The errors that stem from disruption to circadian rhythms (biological processes over a 24-hour period) and mental exhaustion, and can lead to injury (e.g., from improper machine operation), can be considered products of cognitive oversight. This oversight, which is an unintentional failure to interpret events correctly, is at the root of much human error in data centers and can potentially result in not just injury, but a disruption to services.
Currently, 8- to10-hour single-day shifts are most common in the data center industry across all major regions, according to the Uptime Institute Data Center Staffing Survey 2023 (Operators struggle to overcome ongoing staff and skills shortage). There are, however, some geographic variations in the results: while 17% of all respondents report single-day shifts of more than 10 hours, Asia-Pacific leads at 22%. In contrast, respondents from Europe have more than three times as many 5- to 7-hour shifts as respondents from Asia-Pacific, but just over half (13%) report shifts of more than 10 hours.
Policy variations across different regions are clearly a factor in how data center owners and operators choose specific shift lengths for their employees, particularly in relation to night shifts. In Europe, labor laws in several major countries do not allow night shifts to exceed 8 or 10 hours as standard. Exceptions can be made to meet 24/7 staffing requirements, with night shifts extended to 12 hours, as long as employees are compensated with sufficient paid time off work.
These policy restrictions in Europe — along with the survey results indicating that European respondents provide more 5- to 7-hour shifts than respondents from other regions — may indicate that these companies are hiring more part-time employees to make up their staffing shortfall.
Companies in other regions attempting to replicate a similar strategy to reduce shift length face obstacles. Unlike European employees, workers in the US and several Latin-American countries risk losing access to healthcare coverage if their shifts become shorter. In the US there is no statutory obligation for the employer to provide healthcare coverage if employees work less than a 40-hour week. Staff are therefore reluctant to reduce their weekly hours.
Employers can limit long shifts — particularly night shifts (which have higher workplace injury risk) — to 8 hours. While this may appear to be an intuitive solution to avoid performance deterioration, Uptime Institute’s technical consultants advise that any change will not be without friction, and shift length may not even be the primary contributory factor. Some key considerations are:
Complacency and ownership. Shift structure should promote sharing of knowledge, break monotony of routines and help develop a sense of inclusion through rotating shifts. Shift silos, such as staff having a fixed schedule, with some only working at weekends or nights, may create unhealthy attitudes resulting from complacency or a lack of team cohesion.
Meeting staff lifestyle preferences. Despite data suggesting that long shifts are detrimental to performance, it is difficult for some operators to cut back hours. Uptime Institute technical consultants often see a staff preference for 12-hour shifts over several days, for the benefits of both additional overtime pay and extended blocks of time off work.
Relief shifts. Consensus in the industry is that extending shifts to more than 12 hours is ultimately worse for the business than sending employees home. For many operators, however, extending shifts to beyond 12 hours is unavoidable as a means of meeting staffing requirements. In practice, identifying individuals that can handle these extended shift lengths is not easy. It is not just very long shifts that carry the risks associated with fatigue. Staff not being able to rest sufficiently due to covering the shifts of absentee staff is another source of potential exhaustion, even if these shifts are not particularly long.
Long-term impact
Sourcing the appropriate, qualified individual for a relief shift in an understaffed industry is challenging. Typically, companies request employees to clock in on their rest days. This may work well for an employee during a week they are already off work, but it could also force employees to clock back on before they have had sufficient rest between shifts. Adding more staff into the shift rotation may prevent other employees from having to extend shifts or clock in with insufficient rest, but this simply patches over the root of the problem: the absence of staff from their scheduled shifts.
Operators need to monitor absence levels and understand the reasons behind these absence levels. The cumulative long-term impact of working shifts of more than 10 hours increases the risk of developing a range of health conditions, as well as fatigue. Although many data center operators have developed shift schedules to minimize errors, this needs to be balanced with a long-term view of health, work life balance and burn-out.
Planning ahead
Retroactively adjusting shift lengths of established employees could result in low morale and counterintuitively result in higher levels of fatigue as staff adjust to their new schedule changes. Many data center owners and operators, however, are undergoing significant infrastructure expansion, which need to be staffed on a shift rotation that minimizes human error and limits the risks of disruption to service availability. Owners and operators should consider the following recommendations:
Avoid shift lengths of more than 12 hours. Staffing levels and schedules should be defined to minimize the occurrences of abnormally long shifts.
Identify shifts that are not appropriate as relief shifts. Establish a system for ensuring well-rested coverage. Monitor overtime and rest periods between shifts to avoid calling in exhausted staff.
Consider individual employee preferences but remain mindful that shift workers often ignore potential risks to their own job performance and health when requesting their preferred schedule.
The Uptime Intelligence View
While many data center managers take a flexible approach to staffing, relief shifts remain a common source of human error. Employees experiencing long-term effects of extended shift work, in terms of risks to health and performance, may be perpetuating difficulties in filling the required shifts due to increased levels of staff absence. These factors can result in an operational stress of lower-than-ideal staffing levels in many facilities, leaving data center managers with few options to optimize shifts.
https://journal.uptimeinstitute.com/wp-content/uploads/2024/05/Long-shifts-in-data-centers-time-to-reconsider-featured.jpg5391030Rose Weinschenk, Research Associate, Uptime Institute, [email protected]https://journal.uptimeinstitute.com/wp-content/uploads/2022/12/uptime-institute-logo-r_240x88_v2023-with-space.pngRose Weinschenk, Research Associate, Uptime Institute, [email protected]2024-05-22 15:00:002024-05-21 14:17:51Long shifts in data centers — time to reconsider?
Data center management software is evolving — at last
/in Executive, Operations/by Max Smolaks, Research Analyst, [email protected]Despite their role as enablers of technological progress, data center operators have been slow to take advantage of developments in software, connectivity and sensor technologies that can help optimize and automate the running of critical infrastructure.
Most data center owners and operators currently use building management systems (BMS) and / or data center infrastructure management (DCIM) software as their primary interfaces for facility operations. These tools have important roles to play in data center operations but have limited analytics and automation capabilities, and they often do little to improve facility efficiency.
To handle the increasing complexity and scale of modern data centers — and to optimize for efficiency — operations teams need new software tools that can generate value from the data produced by the facilities equipment.
For more than a decade, DCIM software has been positioned as the main platform for organizing and presenting actionable data — but its role has been largely passive. Now, a new school of thought on data center management software is emerging, proposed by data scientists and statisticians. From their point of view, a data center is not a collection of physical components but a complex system of data patterns.
Seen in this way, every device has a known optimal state, and the overall system can be balanced so that as many devices as possible are working in this state. Any deviation would then indicate a fault with equipment, sensors or data.
This approach has been used to identify defective chiller valves, imbalanced computer room air conditioning (CRAC) fans, inefficient uninterruptible power supply (UPS) units and opportunities to lower floor pressure — all discovered in data, without anyone having to visit the facility. The overall system — the entire data center — can also be modelled in this way, so that it can be continuously optimized.
All about the data
Data centers are full of sensors. They can be found inside CRAC and computer room air handler units, chillers, coolant distribution units, UPS systems, power distribution units, generators and switchgear, and many operators install optional temperature, humidity and pressure sensors in their data halls.
These sensors can serve as a source of valuable operational insight — yet the data they produce is rarely analyzed. In most cases, applications of this information are limited to real-time monitoring and basic forecasting.
In recent years, major mechanical and electrical equipment suppliers have added network connectivity to their products as they look to harvest sensor data to understand where and how their equipment is used. This has benefits for both sides: customers can monitor their facilities from any device, anywhere, while suppliers have access to information that can be used in quality control, condition-based or predictive maintenance, and new product design.
The greatest benefits of this trend are yet to be harnessed. Aggregated sensor data can be used to train artificial intelligence (AI) models with a view to automating an increasing number of data center tasks. Data center owners and operators do not have to rely on equipment vendors to deliver this kind of innovation. They can tap into the same equipment data, which can be accessed using industry-standard protocols like SNMP or Modbus.
When combining sensor data with an emerging category of data center optimization tools — many of which rely on machine learning — data center operators can improve their infrastructure efficiency, achieve higher degrees of automation and lower the risk of human error.
The past few years have also spawned new platforms that simplify data manipulation and analysis. These enable larger organizations to develop their own applications that leverage equipment data — including their own machine learning models.
The new wave
Dynamic cooling optimization is the best-understood example of this new, data-centric approach to facilities management. These applications rely on sensor data and machine learning to determine and continually “learn” the relationships between variables such as rack temperatures, cooling equipment settings and overall cooling capacity. The software can then tweak the cooling equipment performance based on minute changes in temperature, enabling the facility to respond to the needs of the IT in near real-time.
Many companies working in this field have close ties to the research community. AI-powered cooling optimization vendor TycheTools was founded by a team from the Technical University of Madrid. The Dutch startup Coolgradient was co-founded by a data scientist and collaborates with several universities. US-based Phaidra has brought together some of the talent that previously published and commercialized cutting-edge research as part of Google’s DeepMind.
Some of the benefits offered by data-centric management software include:
Not all machine learning models require extensive compute resources, rich datasets and long training times. In fact, many of the models used in data centers today are small and relatively simple. Both training and inference can run on general-purpose servers, and it is not always necessary to aggregate data from multiple sites — a model trained locally on a single facility’s data will often be sufficient to deliver the expected results.
New tools bring new challenges
The adoption of data-centric tools for infrastructure management will require owners and operators to recognize the importance of data quality. They will not be able to trust the output of machine learning models if they cannot trust their data — and that means additional work on standardizing and cleaning their operational data stores.
In some cases, data center operators will have to hire analysts and data scientists to work alongside the facilities and IT teams.
Data harvesting at scale will invariably require more networking inside the data center — some of it wireless — and this presents a potentially wider attack surface for cybercriminals. As such, cybersecurity will be an important consideration for any operational AI deployment and a key risk that will need to be continuously managed.
Evolution is inevitable
Uptime Institute has long argued that data center management tools need to evolve toward greater autonomy. The data center maturity model (see Table 1) was first proposed in 2019 and applied specifically to DCIM. Four years later, there is finally the beginning of a shift toward Level 4 and Level 5 autonomy, albeit with a caveat — DCIM software alone will likely never evolve these capabilities. Instead, it will need to be combined with a new generation of data-centric tools.
Table 1. The Data Center Management Maturity Model
Not every organization will, or should, take advantage of Level 4 and Level 5 functionality. These tools will provide an advantage to the operators of modern facilities that have exhausted the list of traditional efficiency measures, such as those achieving PUE values of less than 1.3.
For the rest, being an early adopter will not justify the expense. There are cheaper and easier ways to improve facility efficiency that do not require extensive data standardization efforts or additional skills in data science.
At present, AI and analytics innovation in data center management appears to be driven by startups rather than established software vendors. Few BMS and DCIM developers have integrated machine learning into their core products, and while some companies have features in development, these will take time to reach the market — if they ever leave the lab.
Uptime Intelligence is tracking six early-stage or private companies that use facilities data to create machine learning models and already have products or services on the market. It is likely more will emerge in the coming months and years.
These businesses are creating a new category of software and services that will require new types of interactions with all the moving parts inside the data center, as well as new commercial strategies and new methods of measuring the return on investment. Not all of them will be successful.
The speed of mainstream adoption will depend on how easy these tools will be to implement. Eventually, the industry will arrive at a specific set of processes and policies that focus on benefitting from equipment data.
The Uptime Intelligence View
The design and capabilities of facilities equipment have changed considerably over the past 10 years, and traditional data center management tools have not kept up. A new generation of software from less-established vendors now offers an opportunity to shift the focus from physical infrastructure to data. This introduces new risks — but the benefits are too great to ignore.
Generative AI and global power consumption: high, but not that high
/in Executive, Operations/by Andy Lawrence, Executive Director of Research, Uptime Institute, [email protected]In the past year, Uptime Intelligence has been asked more questions about generative AI and its impact on the data center sector than any other topic. The questions come from enterprise and colocation operators, suppliers of a wide variety of equipment and services, regulators and the media.
Most of the questions concern power consumption. The compute clusters — necessary for the efficient creation and use of generative AI models — are enormously power-hungry, creating a surge in projected demand for capacity and, for operators, challenges in power distribution and cooling in the data center.
The questions about power typically fall into one of three groups. The first is centered around, “What does generative AI mean for density, power distribution and cooling in the data center?”
Uptime Intelligence’s view is that the claims of density distress are exaggerated. While the AI systems for training generative AI models, equipped with Nvidia accelerators, are much denser than usual, they are not extreme and can be managed by spreading them out into more cabinets. Further Uptime Intelligence reports will cover this aspect since it is not the focus here.
Also, generative AI has an indirect effect on most data center operators in the form of added pressures on the supply chain. With lead times on some key equipment, such as engine generators, transformers and power distribution units, already abnormally long, unforeseen demand for even more capacity by generative AI will certainly not help. The issues of density and forms of indirect impact were initially addressed in the Uptime Intelligence report Hunger for AI will have limited impact on most operators.
The second set of questions relates to the availability of power in a given region or grid — and especially of low-carbon energy. This is also a critical, practical issue that operators and utilities are trying to understand. The largest clusters for training large generative AI models comprise many hundreds of server nodes and draw several megawatts of power at full load. However, the issues are largely localized, and they are also not the focus here.
Generative AI and global power
Instead, the focus of this report is the third typical question, “How much global power will AI use or require?” While in immediate practical terms this is not a key concern for most data center operators, the headline numbers will shape media coverage and public opinion, which ultimately will drive regulatory action.
However, some of the numbers on AI power circulating in the press and at conferences, cited by key influencers and other parties, are extremely high. If accurate, these figures suggest major infrastructural and regulatory challenges ahead — however, any unduly high forecasts may prompt regulators to overreact.
Among the forecasts at the higher end is from Schneider Electric’s respected research team, which estimated AI power demand at 4 gigawatts (GW), equivalent to 35 terawatt-hours (TWh) if annualized, in 2023, rising to around 15 GW (131 TWh if annualized) in 2028 (see Schneider White paper 110: The AI Disruption: Challenges and Guidance for Data Center Design). Most likely, these figures include all AI workloads and not only new generative models.
And Alex de Vries, of digital trends platform Digiconomist and whose calculations of Bitcoin energy use have been influential, has estimated AI workload use at 85 TWh to 134 TWh by 2027. These figures suggest AI could add 30% to 50% or more to global data center power demand over the next few years (see below).
There are two reasons why Uptime Intelligence considers these scenarios overly bullish. First, estimates on power for all AI workloads are problematic for both taxonomy (what is AI) and the improbability of tracking it meaningfully. Also, most forms of AI are already accounted for in capacity planning, despite generative AI being unexpected. Second, projections that span beyond a 12- to 18-month horizon carry high uncertainty.
Some of the numbers cited above imply an approximately thousand-fold increase in AI compute capacity in 3 to 4 years from the first quarter of 2024, when accounting for hardware and software technology evolution. That is not only unprecedented but also has weak business fundamentals when considering the hundreds of billions of dollars it would take to build all that AI infrastructure.
Uptime Intelligence takes a more conservative view with its estimates — but these estimates still indicate rapidly escalating energy use by new large generative AI models.
To reach our numbers, we have estimated the current shipments and installed base of Nvidia-based systems through to the first quarter of 2025 and the likely power consumption associated with their use. Systems based on Nvidia’s data center accelerators, derived from GPUs, dominate the generative AI model accelerator market and will continue to do so until at least mid-2025 due to an entrenched advantage in the software toolchain.
We have considered a range of technical and market factors to calculate the power requirements of generative AI infrastructure: workload profiles (workload activity, utilization and load concurrency in a cluster), the shifting balance between training and inferencing, and the average PUE of the data center housing the generative AI systems.
This data supports our earlier stated view that the initial impact of generative AI is limited, beyond a few dozen large sites. For the first quarter of 2024, we estimate the annualized power use by Nvidia systems installed to be around 5.8 TWh. This figure, however, will rise rapidly if Nvidia meets its forecast sales and shipment targets. By the first quarter of 2025, the generative AI infrastructure in place could account for 21.9 TWh of annual energy consumption.
We expect these numbers to shift as new information emerges, but they are indicative. To put these numbers into perspective, the total global data center energy use has been variously estimated at between 200 TWh and 450 TWh per year in the periods from 2020 to 2022. (The methodologies and terms of various studies vary widely and suggest that data centers use between 1% and 4% of all consumed electricity.) By taking a middle figure of 300 TWh for the annual global data center power consumption, Uptime Intelligence puts generative AI annualized energy at around 2.3% of the total grid power consumption by data centers in the first quarter of 2024. However, this could reach 7.3% by the first quarter of 2025.
Outlook
These numbers indicate that generative AI’s power use is not, currently, disruptively impactful, given the data center sector’s explosive growth in recent years. Generative AI’s share of power consumption relative to its footprint is outsized given its likely very high utilization. A share of data center capacity by power will fall in the low-to-mid single digits even by the end of 2025.
However, it is nevertheless a dramatic increase and suggests that generative AI could account for a much larger percentage of overall data center power use in the years ahead. Inevitably, the surge needs to slow down, also in great part because newer AI systems built around vastly more efficient accelerators will displace the install base en masse rather than adding net new infrastructure.
While Uptime Intelligence believes some of the estimates of generative AI power use (and data center power use) to be too high, the sharp uptick — and the concentration of demand in certain regions — will still be high enough to attract and stimulate regulatory attention. Uptime Intelligence will continue to analyze the development of AI infrastructure and its impact on the data center industry.
Andy Lawrence, Executive Research Director [email protected]
Daniel Bizo, Research Director dbizo@uptimeinstitute.com
Understanding how server power management works
/in Design, Executive, Operations/by Daniel Bizo, Research Director, Uptime Institute Intelligence, [email protected]Uptime Intelligence regularly addresses IT infrastructure efficiency, particularly servers, in our reports on data center energy performance and sustainability. Without active contribution from IT operations, facility operations alone will not be able to meet future energy and sustainability demands on data center infrastructure. Purchases of renewable energy and renewable energy certificates will become increasingly — and, in many locations, prohibitively — expensive as demand outstrips supply, making the energy wasted by IT even more costly.
The power efficiency of a server fleet, that is, how much work servers perform for the energy they use, is influenced by multiple factors. Hardware features receive the most attention from IT buyers: the server’s technology generation, the configuration of the system and the selection of power supply or fan settings. The single most significant factor that affects server efficiency, however, is the level at which the servers are typically utilized; a seemingly obvious consideration — and enough for regulators to include it as a reporting requirement in the EU’s new Energy Efficiency Directive (see EED comes into force, creating an enormous task for the industry). Even so, the process of sourcing the correct utilization data for the purposes of power efficiency calculations (as opposed to capacity planning) remains arguably misunderstood (see Tools to watch and improve power use by IT are underused).
The primacy of server utilization in data center efficiency has increased in recent years. The latest server platforms are only able to deliver major gains in energy performance when put to heavy-duty work — either by carrying a larger software payload through workload consolidation, or by running scalable, large applications. If these conditions are not met, running lighter or bursty workloads on today’s servers (regardless of whether based on Intel or AMD chips) will deliver only a marginal, if any, improvement in the power efficiency compared with many of the supposedly outdated servers that are five to seven years old (see Server efficiency increases again — but so do the caveats).
Cycles of a processor’s sleep
This leads into the key discussion point of this report: the importance of taking advantage of dynamic energy saving features. Settings for power and performance management of servers are often an overlooked — and underused — lever in improving power efficiency. Server power management techniques affect power use and overall system efficiency significantly. This effect is even more pronounced for systems that are only lightly loaded or spend much of their time doing little work: for example, servers that run enterprise applications.
The reduction in server power demand resulting from power management can be substantial. In July 2023 Uptime Intelligence published a report discussing data (although sparse) that indicates 10% to 20% reductions in energy use from enabling certain power-saving modes in modern servers, with only a marginal performance penalty when running a Java-based business logic (see The strong case for power management). Energy efficiency gains will depend on the type of processor and hardware configuration, but we consider the results indicative for most servers. Despite this, our research indicates that many, if not most, IT operations do not use power management features.
So, what are these power management settings? Server power management settings are governed by the firmware statically (what modes are enabled upon system start up) and dynamically by the operating system or hypervisor once running through the Advanced Configuration and Power Interface.
There are many components in a server that may have power management features, enabling them to run slower or power off. Operating systems also have their own software mechanisms, such as suspending their operation and saving the machine state to central memory or the storage system.
But in servers, which tend to be always powered on, it is the processors’ power management modes that dictate most of the energy gains. Modern processors have sophisticated power management features for idling, that is, when the processor does not execute code. These are represented by various levels of C-states (the C stands for CPU) denoted by numbers, such as C1 and C2 (with C0 being the fully active state).
The number of these states has expanded over time as chip architects introduce new, more advanced power-saving features to help processors reduce their energy use when doing no work. The chief benefit of these techniques is to minimize leakage currents that would otherwise increasingly permeate modern processor silicon.
The higher the C-state number, the more of its circuitry the CPU sends to various states of sleep. In summary:
Skipped numbers, such as C2 and C5, are incremental, transitionary processor power states between main states. Not all these C-states are available on all processor architectures.
A good sleep in a millisecond
The levels of C-states and understanding them matters because they largely define the cost in performance and the benefit in power. The measured results of 10% to 20% reduction in energy use when enabling certain power management features, as discussed earlier, have allowed the server processor (an AMD model) to enter processor power states up to C6. These sleep states save power even when the server, on a human level of perception, is processing database transactions and responding to queries.
This is because processors operate on a timescale measured in nanoseconds, while software-level requests between commands can take milliseconds even on a busy machine. This is a factor of one million difference: milliseconds between work assignments represent millions of processor cycles waiting. For modern server processors, some of the many cores may often have no work to do for a second or more, which is an eternity on the processor’s time scale. On a human scale, a comparable time would be several years of inactivity.
However, there is a cost associated with the processor cores going to sleep. Entering ever deeper sleep states across processor cores or entire chips can take thousands of cycles, and as many as tens of thousands of cycles to wake up and reinstate operation. This added latency to respond to wake-up requests is what shows up as a loss of performance in measurements. In the reference measurement running the Java-based business logic, this is in the 5% to 6% range — arguably a small price to pay.
Workloads will vary greatly in the size of the performance penalty introduced by this added latency. Crucially, they will differ even more in how costly the lost application performance is for the business — high-frequency trading or processing of high volumes of mission-critical online transactions are areas where any loss of performance is unacceptable. Another area may include storage servers with heavy demand for handling random read-write operations at low latency. But a vast array of applications will not see material change to the quality of service.
Using server power management is not a binary decision either. IT buyers can also calibrate the depth of sleep they enable for the server processor (and other components) to enter. Limiting it to C3 or C1E may deliver better trade-offs. Many servers, however, are not running performance-critical applications and spend most of their time doing no work — even if it seems that they are often, by human standards, called upon. For servers that are often idle, the energy saved can be in the 20% to 40% range, which can amount to tens of watts for every lightly loaded or idle server.
Optimizing server energy performance does not stop with C-states. Performance-governing features (setting performance levels when the processor is actively working), known as P-states, offer another set of possibilities to find better trade-offs between power and performance. Rather than minimizing waste when the processor idles, P-states direct how much power should be expended on getting the work done. Future reports will introduce P-states for a more complete view of server power and performance management for IT infrastructure operators that are looking for further options in meeting their efficiency and sustainability objectives.
The Uptime Intelligence View
A server processor’s power management is a seemingly minute function buried under layers of technical details of an infrastructure. Still, its role in the overall energy performance of a data center infrastructure will be outsized for many organizations. In the near term, blanket policies (or simply IT administrator habits) of keeping server power management features switched off will inevitably be challenged by internal stakeholders in pursuit of cost efficiencies and better sustainability credentials; or, possibly in the longer term, by regulators catching on to technicalities and industry practices. Technical organizations at enterprises and IT service providers will want to map out server power management opportunities ahead of time.
EU battery regulations: what do the new rules mean?
/in Executive, Operations/by Rosa Lawrence, Research Associate, Uptime Institute, [email protected]The European Green Deal, a set of policy initiatives approved in 2020, aims for a sustainable and competitive economy with net-zero greenhouse gas emissions by 2050. Along with the legislation driving this transition, such as the Energy Efficiency Directive recast (see EED comes into force, creating an enormous task for the industry), the strategy will require much higher rates of electrification in transport and industries to displace the use of fossil fuels. This aim is combined with the policy objective of adding large amounts of renewable power generation capacity.
However, this direction is leading toward a new environmental challenge: increased sales of electric vehicles (EVs) will create future end-of-life battery problems, drawing attention to the unresolved issue of reusing and recycling large battery banks. Many industrial stationary applications are also seeing a strong take-up of EV-type (mostly lithium-ion) batteries for a variety of reasons. These include the displacement of valve-regulated lead-acid (VRLA) batteries, which are highly recycled, new energy storage installations for grid demand-response schemes and the elimination of standby engine generators.
Until now, this area has been governed by the 2006 Battery Directive (2006/66/EC). However, this directive is being replaced by Regulation 2023/1542 — the main objective of which is to create an updated set of rules to ensure high sustainability standards regardless of the battery chemistry. As an EU Regulation, it applies automatically to all member states without implementation in national laws.
The new regulations address aspects such as carbon footprint, recycled content, safety, labeling and end-of-life management. Industrial customers, including data center suppliers and their customers, will need to start reporting by mid-2025 — although this date may yet change. Although these rules only apply to members of the EU, its standards and laws are often replicated in other countries.
The 2006 Battery Directive was put in place to mitigate the environmental impact of battery production and disposal of waste batteries. It introduced recycling and treatment targets, restrictions on some hazardous substances, labeling requirements (for hazardous substances and instructions for proper disposal), reporting obligations and extended producer responsibility.
Lead-acid versus Li-ion recycling: the facts
The most common batteries used in uninterruptible power supply (UPS) systems are VRLA and Li-ion batteries (of which several sub-types exist). Lead-acid batteries have the highest collection and recycling rates. In the EU, the recycling rate for automotive starter batteries is 99% and more than 90% of the lead is recovered. The figure for stationary applications, which includes data centers, is likely to be similarly high due to the stringent regulations and economic incentives to recycle.
In 2021, all EU member states met the target recycling rate of 65% by weight for lead-acid batteries (both automotive and non-automotive).
The recycling process of lead-acid batteries consists of draining the electrolyte, opening the casing and separating the materials. The lead plates are then smelted to obtain molten lead, which is purified and refined before being cast into ingots for reuse. The plastic components are also recycled. Lead-acid batteries have few components and contain approximately 70% lead, which means that it is an efficient process. It is also profitable because recycled lead can be used in the manufacture of new batteries and is recyclable at relatively low temperatures, which require less energy.
For Li-ion batteries, the view is more complicated. Currently, it is cheaper to mine the metals to make new Li-ion cells (regardless of Li-ion chemistry) than it is to recycle used batteries. As a result, the much higher environmental footprint of the supply chain to produce Li-ion cells has so far been an externality and not priced into the cost of making Li-ion batteries. This factors heavily into its current market advantage in price-performance compared to other, environmentally less damaging cell chemistries.
The two most common processes for recycling spent Li-ion batteries are pyrometallurgy and hydrometallurgy. During both processes, the batteries are discharged and dismantled. They are then either smelted at high temperatures or leached to recover high-value cathode materials, such as cobalt, nickel and copper. During pyrometallurgical smelting, the lithium is lost in the furnace, making the process for lithium not sustainable. Future recycling processes for Li-ion cells are in development and will likely involve a combination of heat treatment and leaching in various concentrations of acid to maximize recovery of valuable metals.
These processes are energy-intensive and therefore expensive to operate, as well as producing gaseous pollutants and industrial waste. According to some battery vendors, the costs involved with the transport and recycling of Li-ion batteries, which is considered a hazardous waste, are substantial.
This is a major reason behind the current expectation that Li-ion battery packs will find secondary and even tertiary uses rather than being recycled for their raw materials. Compared with lead-acid batteries, this is possible due to Li-ion cells’ higher endurance and shelf life. The new battery regulation actively considers this scenario with the introduction of battery passports.
What are the new rules?
The phased implementation of the rules (Regulation 2023/1542) begins in July 2024 and regulates the carbon footprint, recycled content of new batteries, labeling and the introduction of an online battery information system. The new battery regulation controls all battery chemistries, with rules varying by battery category, for example, EV, industrial and portable. Recycling targets differ between chemistries, with specific targets for the recovery of cobalt, lead, lithium and nickel. Figure 1 sets out the major milestones of the regulatory rollout for EU member states, as prescribed in the regulation.
Figure 1. Timeline of the EU’s battery regulation
New batteries put to market will be subject to mandatory minimum levels of recycled content requirements. From 2030, batteries will need to contain a minimum recycled content of 12% for cobalt, 4% for lithium, 4% for nickel and 85% for lead. By 2035, these thresholds will increase to 20% cobalt, 10% lithium, 12% nickel and 85% lead.
Alongside these requirements, there are also recycling efficiency and material recovery targets for end-of-life batteries. By the end of 2030, used batteries will have a recycling target by weight of 80% for lead-acid and 70% for Li-ion. The material recovery target is 95% for cobalt, copper, lead and nickel and 70% for lithium.
What does this mean for data center operators?
The new rules will mostly affect battery producers; however, some responsibilities will fall on end users, such as data center operators. These include performing due diligence to verify vendor (product carbon footprint) and recycler claims (environmental credentials of the recycling process), as well as maintaining up-to-date battery passports during the use of the products.
From August 18, 2025, battery suppliers and data center operators (with some exceptions), will have a legal requirement to adopt a battery due diligence policy covering the social and environmental risks. This policy will need to include annual reports on the risks in the supply chain and the steps taken to manage them, transparency on sourcing raw and recycled materials, and structuring an internal management system to support it. The due diligence policy should be verified by a third party and communicated to suppliers and the public via annual reports.
There are exceptions for suppliers and buyers of batteries with an annual net turnover of less than €40 million ($43 million) and those using batteries that have undergone preparation for reuse, repurposing, or remanufacturing before being placed on the market.
The introduction of battery passports in 2026 will affect both suppliers and end users, including data center operators. Battery producers will be required to report on properties such as the chemistry, carbon footprint and recycled content, while it will be the end user’s responsibility to update the battery passport on the state of its health throughout its life. This will likely include maintenance entries but also any substantial events or accidents, such as deviations from standard operating environmental conditions, overcharging or deep discharging, and other impacts.
Battery passports will be key to enabling a second-hand market for batteries that are appropriate for reuse. Alongside the labeling of batteries, this may lead to improvements in Li-ion battery recycling by simplifying the separation process by sorting batteries by chemistry before the start of the recycling process. However, the technical implementation of the battery passport has not been stipulated in the new regulation and will be left to future cooperation between EU member states.
The regulation states that producers shall cover the necessary costs incurred by the collection and recycling of waste batteries. Lead-acid batteries have an inherent economic value at the end of their useful lives, which guarantees incentives for both buyers and sellers to promote recycling. Li-ion batteries, however, currently incur substantial costs to pay for recycling. According to battery makers, this can amount to as much as the price of new Li-ion batteries. Another potentially overlooked detail is the cost of transporting large amounts of Li-ion batteries that are classified as hazardous and pose a fire risk.
It is possible (or even likely) that end users may see a marked increase in the price of Li-ion batteries — the selling price will need to cover the recycling costs as a result of this extended producer responsibility. Data center operators, with hundreds of batteries in their UPS systems, may have to set up databases and processes to ensure they are reliably tracked.
The regulation stipulates that the ultimate responsibility for managing end-of-life batteries will fall on suppliers — in practical terms: taking batteries from end-users for no charge and ensuring the batteries are reused or recycled. However, end users should also exercise due diligence, even when they are not legally required to at present. Data center owners and operators should be aware that battery manufacturers may overstate their recycling capabilities. This is particularly relevant given the long lifespan of industrial Li-ion batteries, which can remain in use for 10 to 15 years or longer. Companies placing such batteries on the market may not have adequate recycling plans in place.
End users should also consider what technology is used in the recycling facilities, and even verify the existence of these facilities by checking their reported location using satellite imagery and public records. Inevitably, some battery vendors will go out of business, leaving end users with a potentially costly liability.
Additionally, when considering Scope 3 emissions reporting (in accordance with the Greenhouse Gas Protocol), the new recycling obligations add complexity. End users will need to consider carbon emissions from transportation and recycling processes. While some may argue that recycling offsets carbon emissions compared with manufacturing new batteries from raw materials, this claim may not withstand closer scrutiny.
Where could these rules go in the future?
The 2023 battery regulation provides a timeline of implementation, but the rules will need further clarification when they start to come into effect. Potential secondary legislation could involve standardizing calculations for carbon footprint, recycling efficiency and material recovery. More clarification may follow for the implementation and functioning of the electronic exchange system as well as specifications for the content and accessibility of battery passports.
Additionally, EU member states may choose to enact further requirements (as long as they are not in conflict with EU law) to stimulate investments in research and development related to carbon footprint reduction and sustainability in battery production and recycling.
The Uptime Intelligence View
The new EU Battery Regulation is primarily a response to the mass-market adoption of EVs, however, it also covers industrial stationary applications, such as mission-critical power systems.
The EU’s objective is to ensure that huge quantities of new batteries will not simply end up as hazardous waste at the end of their lives but will either find new uses or be recycled to make new battery cells. It will also level the playing field with lead-acid batteries and other, more readily recyclable chemistries.
Rosa Lawrence, Research Associate, Uptime Institute, [email protected]
Daniel Bizo, Research Director, Uptime Institute, [email protected]
Colocation and public cloud growth masks enterprise expansion
/in Executive, Operations/by Douglas Donnellan, Senior Research Associate, Uptime Institute, [email protected]The colocation and public cloud sectors of the digital infrastructure industry continue to make headlines, with many organizations planning large-scale capacity expansion to meet rising demand. However, there is also a less public expansion underway — enterprises operators, for the third successive year, say they are going to invest in more data center capacity in 2024.
Results from the Uptime Institute Capacity Trends Survey 2023 reveal that 64% of enterprise operators are growing their data center capacity — a six percentage-point uptick from two years earlier in 2021 (see Figure 1). Notably, one in five organizations in this group say they are expanding by more than 20% annually. This scale of expansion is difficult to implement without significant investment.
Figure 1. Enterprise data center capacity shows strong growth
Some suppliers — and some operators — may be surprised by this rate of growth since it follows a decade-long period in which enterprise data centers were often dismissed as expensive, inflexible and outdated by many executives. However, Uptime’s data, which has been consistent over the past three years and is backed by many conversations, suggests that enterprise investment is strong.
What is driving this growth in enterprise data center capacity? Partly, it is simply a demand for more digital services. However, companies also say that they are investing to enhance the resiliency of their data centers (45% of survey respondents) and to support a hybrid cloud strategy (37%). Moving to cloud architectures often requires an increase in data center capacity, especially if an organization is developing distributed resiliency architectures.
Cost may also be a factor that is driving enterprise investment. According to separate Uptime research, of those enterprise operators that compared the cost of provisioning workloads on-premises versus off-premises, most report that corporate data centers are less expensive than using colocation (56%, n=154) or public cloud (51%, n=151). This is particularly true for enterprises that have already made significant investments to expand capacity.
One looming problem for enterprises — and indeed for colocation companies — is the widely forecast increase in rack power density in the coming years. To accommodate this, new investments in cooling and power distribution will be required.
Four-fifths (82%) of enterprises say that they are expecting more demand for higher power densities in the next two to three years, but more than one-third (36%) say that they cannot accommodate this demand with their existing infrastructure (see Figure 2). As a result, many workloads with higher density demands will, as Uptime expects, be outsourced to third parties that have the requisite power and cooling infrastructure.
Figure 2. Many need new investment to meet expected power densities
In spite of the increased enterprise sector spending, the trend towards greater outsourcing to colocation and cloud companies is expected to remain strong (see The majority of enterprise IT is now off-premises). For example, colocation companies report more growth than their enterprise counterparts by 15 percentage points (79%, n=130), with twice as many reporting annual growth rates of more than 20% (40%, n=130).
Taken together, Uptime’s survey data shows that chief information officers are investing in cloud, hosting, colocation and enterprise data centers. While more workloads may be outsourced, the enterprise data center will most likely continue to grow and evolve. Large companies with complex mission-critical workloads, especially those that are heavily regulated, will most likely maintain on-premises sites.
However, as colocation and public cloud providers expand the depth of their services in response to the industry’s staffing, regulatory and supply chain challenges, enterprises will increasingly integrate these resources over the next decade.
The Uptime Intelligence View
Enterprise data centers have been characterized as being in decline over the past five years, especially in the context of the significant, double-digit annual growth of large colocation and public cloud organizations. But Uptime survey data has consistently shown investment in the sector. Although owning and operating data centers may not feature in the strategies employed by many large and especially newer organizations, enterprise facilities will likely remain essential to businesses beyond the medium term.
Long shifts in data centers — time to reconsider?
/in Executive, Operations/by Rose Weinschenk, Research Associate, Uptime Institute, [email protected]Human error has been — and remains to be — a major cause of outages in data centers. Uptime Intelligence’s research shows that about four in 10 operators have had a major outage in the past three years in which human error played a role (Annual outage analysis 2023). Half of these respondents said errors were made because staff failed to follow the correct procedures.
Thorough training, regular practice in equipment testing and work experience all help to reduce these errors — particularly in an emergency when a prompt reaction is crucial. An often underappreciated factor is the importance of mental performance and the effects of fatigue.
The relationship between shift length, fatigue and human error is well documented, but less clear is how the data center industry can define shifts that help minimize human error. The recommended best practices for other industries do not always translate into the data center world, where 24/7 service availability is the standard. Additionally, data center owners and operators wanting to optimize shift length to limit fatigue need to navigate employee preferences and region-specific constraints.
What the research says
Studies indicate there is a tipping point after which the performance of most staff deteriorates. Researchers at the Chinese University of Hong Kong Department of Systems Engineering and Engineering Management analyzed 241 papers on the relationship between shift length and occupational health and found that individuals working more than 10-hour shifts are significantly more likely to experience fatigue. A similar review from the Finnish Institute of Occupational Health shows the risk of workplace injury due to fatigue-related accidents across a range of industries is 15% higher in 10-hour shifts than 8-hour shifts, and jumps to 38% higher at 12 hours.
The errors that stem from disruption to circadian rhythms (biological processes over a 24-hour period) and mental exhaustion, and can lead to injury (e.g., from improper machine operation), can be considered products of cognitive oversight. This oversight, which is an unintentional failure to interpret events correctly, is at the root of much human error in data centers and can potentially result in not just injury, but a disruption to services.
Currently, 8- to10-hour single-day shifts are most common in the data center industry across all major regions, according to the Uptime Institute Data Center Staffing Survey 2023 (Operators struggle to overcome ongoing staff and skills shortage). There are, however, some geographic variations in the results: while 17% of all respondents report single-day shifts of more than 10 hours, Asia-Pacific leads at 22%. In contrast, respondents from Europe have more than three times as many 5- to 7-hour shifts as respondents from Asia-Pacific, but just over half (13%) report shifts of more than 10 hours.
Policy variations across different regions are clearly a factor in how data center owners and operators choose specific shift lengths for their employees, particularly in relation to night shifts. In Europe, labor laws in several major countries do not allow night shifts to exceed 8 or 10 hours as standard. Exceptions can be made to meet 24/7 staffing requirements, with night shifts extended to 12 hours, as long as employees are compensated with sufficient paid time off work.
These policy restrictions in Europe — along with the survey results indicating that European respondents provide more 5- to 7-hour shifts than respondents from other regions — may indicate that these companies are hiring more part-time employees to make up their staffing shortfall.
Companies in other regions attempting to replicate a similar strategy to reduce shift length face obstacles. Unlike European employees, workers in the US and several Latin-American countries risk losing access to healthcare coverage if their shifts become shorter. In the US there is no statutory obligation for the employer to provide healthcare coverage if employees work less than a 40-hour week. Staff are therefore reluctant to reduce their weekly hours.
Employers can limit long shifts — particularly night shifts (which have higher workplace injury risk) — to 8 hours. While this may appear to be an intuitive solution to avoid performance deterioration, Uptime Institute’s technical consultants advise that any change will not be without friction, and shift length may not even be the primary contributory factor. Some key considerations are:
Long-term impact
Sourcing the appropriate, qualified individual for a relief shift in an understaffed industry is challenging. Typically, companies request employees to clock in on their rest days. This may work well for an employee during a week they are already off work, but it could also force employees to clock back on before they have had sufficient rest between shifts. Adding more staff into the shift rotation may prevent other employees from having to extend shifts or clock in with insufficient rest, but this simply patches over the root of the problem: the absence of staff from their scheduled shifts.
Operators need to monitor absence levels and understand the reasons behind these absence levels. The cumulative long-term impact of working shifts of more than 10 hours increases the risk of developing a range of health conditions, as well as fatigue. Although many data center operators have developed shift schedules to minimize errors, this needs to be balanced with a long-term view of health, work life balance and burn-out.
Planning ahead
Retroactively adjusting shift lengths of established employees could result in low morale and counterintuitively result in higher levels of fatigue as staff adjust to their new schedule changes. Many data center owners and operators, however, are undergoing significant infrastructure expansion, which need to be staffed on a shift rotation that minimizes human error and limits the risks of disruption to service availability. Owners and operators should consider the following recommendations:
The Uptime Intelligence View
While many data center managers take a flexible approach to staffing, relief shifts remain a common source of human error. Employees experiencing long-term effects of extended shift work, in terms of risks to health and performance, may be perpetuating difficulties in filling the required shifts due to increased levels of staff absence. These factors can result in an operational stress of lower-than-ideal staffing levels in many facilities, leaving data center managers with few options to optimize shifts.