According to analysis from the International Energy Agency (IEA), “Global electricity demand from data centers is set to more than double during the next five years, consuming as much electricity by 2030 as the whole of Japan does today.” This is just one of many alarming predictions that bring the power crisis facing data centers and, by extension, everyone else, into sharp focus.

Central to the data center power crisis is the ongoing, exponential rise in data volumes, largely driven by digital transformation and, increasingly, AI. Indeed, the amount of electricity required by AI systems could equate to nearly half of total global data center power consumption by the end of this year, if predictions are correct.

Another major part of the problem is that many data centers still rely on legacy infrastructure, including older servers and storage media, which operate continuously at high power levels regardless of workload. Add to this widespread power distribution inefficiencies, backup systems, and the physical size and location of data centers, and the result is a substantial increase in overall energy consumption, including the need for additional cooling, lighting, and environmental controls.

Market forces are also at play, however, and despite these extremely serious issues, the data center sector continues to thrive. As the World Economic Forum puts it, data centers are “not just digital infrastructure, but also high-value investment assets.” In monetary terms, this translates to an industry worth $242.72 billion today, but projected to more than double to more than $584bn by 2032.

The data hoarding imperative

Compounding the power capacity challenge is the fact that organizations are holding onto more data for longer, and often indefinitely. This trend has shifted data center storage profiles, and by 2030, roughly 70 percent of stored data will be classified as ‘cold’, meaning it will be rarely accessed but still required. Meanwhile, only about 30 percent will be ‘hot’ data that demands fast access. Storing this ‘standby’ data on power-hungry devices creates a substantial and growing energy burden that directly contributes to rising electricity consumption in data centers.

Part of the challenge is that, quite understandably, the value of data is measured not only in its current use but also in its potential future applications. This leads to a ‘don’t delete anything’ mindset, where businesses err on the side of caution, preserving data in case it becomes useful down the line.

Addressing this data hoarding imperative depends on balancing the desire to extract maximum value from data with the cost and sustainability implications of storing it indefinitely. Without strategic changes, this culture will continue to intensify power consumption pressures on data center infrastructure.

Making the case for storage transformation

So, what should organizations be doing instead? As part of a concerted effort to address the power consumption crisis, data center leaders should begin by auditing data usage workflows and categorizing their storage needs. Armed with this insight, investing in hybrid architectures that intelligently allocate data across efficient workflows based on access and criticality will significantly improve sustainability performance.

IT teams are then also in a much better position to migrate standby data to energy-efficient storage solutions such as tape, delivering substantial cost and power savings in the process. The scope for lowering energy consumption is significant and, based on the combined installed base of disk and tape storage worldwide, for example, converting around one-third of disk storage to tape has the potential to reduce energy consumption by hundreds of millions of megawatt-hours annually.

Beyond financial benefits, organizations face growing pressure from regulators and various stakeholders to reduce energy consumption. Adopting low-power tape storage aligns with these objectives while also supporting key objectives around business continuity and data security.

Looking ahead

There’s no doubt that data center power consumption will remain a critical challenge as data volumes continue to grow. Sustainability requires ongoing innovation in both infrastructure design and operational practices, particularly given the extremely long lead times required to build generating capacity at scale.

Emerging technologies, including advanced cooling methods, improved resource management, and more energy-efficient hardware, offer promising avenues for reducing power consumption. Somewhat ironically, in this context, AI and automation can also play a role in optimizing workloads, dynamically allocating resources, enhancing cooling efficiency, and enabling data centers to adapt to fluctuating demand.

Storage innovation will also play a key role. Continued development of tape technology, alongside more intelligent hybrid storage solutions, will help balance capacity, cost, and power efficiency, while data classification and automated tiering will become essential for managing expanding data sets in a sustainable manner.

At the same time, regulatory pressures and corporate ESG commitments are certain to build pressure on organizations to balance their tech-led ambitions with credible sustainability efforts. This is critical if the data center industry is to continue its success story.