For years, data-center owners found themselves in a complex relationship with tech giants. These large companies, including Amazon, Microsoft, and Google, had an insatiable demand for data-center space. However, their sheer size and bargaining power allowed them to dictate terms, often driving down rental prices and stunting growth for data-center landlords like Equinix. This dynamic, which saw rents per kilowatt decrease on average by 3% annually between 2013 and 2021, left landlords struggling to maximize returns despite rising demand. But the tides are turning, as a new wave of technological advancements, especially in artificial intelligence (AI), is reshaping the landscape.

Rising Demand and Diminishing Supply: A Landlord’s Market

The data-center market has seen a significant shift in recent years. A key indicator of this change is the sharp decline in vacancy rates across North America's primary data-center markets, hitting a record low of 2.8% in the first half of 2024. This scarcity is driving up rental prices, particularly for hyperscale tech tenants, with rents increasing by 19% over the same period.

The catalyst for this surge in demand is the AI boom. Tech giants like Alphabet, Amazon, Microsoft, and Meta have collectively funneled over $50 billion into capital expenditures in just the second quarter of 2024, a significant portion of which is directed towards AI-related initiatives. As companies rush to train their AI models and enhance their computing capabilities, the need for robust, high-capacity data centers has skyrocketed.

Equinix, one of the world’s largest data-center operators, is at the forefront of this transformation. The company’s xScale business, which caters specifically to hyperscale tech clients, has recently secured land and power for a massive data-center campus in Atlanta. This move is a direct response to the AI-driven demand, which is predominantly seen in Equinix's xScale division. However, smaller corporate and government clients, driven by security concerns or regulatory requirements, are also contributing to the demand, preferring to lease data-center capacity directly rather than relying on the public cloud.

The Financial Upswing: Equinix and the Broader Market

The financial impact of these trends is becoming increasingly apparent. Equinix recently raised its full-year profit guidance after reporting a 17% increase in adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) in the second quarter compared to the previous year. Despite some turbulence earlier in the year, when its stock fell from record highs due to a short-seller report alleging accounting mispractices, the company’s stock has still managed to post a modest 1.7% increase so far this year. An independent review found no wrongdoing, and while the Justice Department continues its investigation, Equinix remains a solid performer in the sector.

This positive trajectory is not isolated to Equinix. The broader data-center market is also seeing increased investment and development. Blackstone, a major investment firm, announced a prospective development pipeline of data centers valued at more than $70 billion. Meanwhile, Amazon has committed to spending $100 billion on data centers over the next decade. Despite the massive capital inflows, the market is unlikely to face an oversupply anytime soon. According to Kristina Metzger, vice chairman at CBRE Data Center Capital Markets, 80% of under-construction data centers in North America are already preleased, mostly to investment-grade tenants on 10-to-15-year contracts.

Construction Delays and Power Supply Issues

However, the data-center market is not without its challenges. The construction of new facilities is being hampered by supply-chain bottlenecks, particularly for critical components such as generators, chillers, and transformers. Additionally, power utilities are struggling to keep pace with the rapid expansion needs of data-center operators. As a result, the timeline for building new data centers has significantly lengthened. A facility that might have taken three years to complete in 2022 could now take up to seven years in particularly constrained markets.

These delays, while problematic, also present opportunities for existing data-center landlords. With new supply coming online at a slower pace, the value of existing, operational data centers is likely to increase further. This environment favors landlords who can now push for higher rents, capitalizing on the scarcity of available space and the high demand from tech companies eager to expand their AI capabilities.

Enhancing Data-Center Infrastructure

As the demand for data centers surges, driven largely by AI and high-performance computing, the infrastructure that supports these facilities is undergoing significant evolution. A key aspect of this transformation is the increasing deployment of GPUs (Graphics Processing Units) in data centers, which are critical for handling the massive computational needs associated with AI workloads.

The Rise of High-Density Computing and GPUs

One of the most prominent trends in data center infrastructure is the shift toward higher rack densities. As AI and machine learning applications become more prevalent, they require increasingly powerful computational resources, which in turn demand higher density setups within data centers. This is where GPUs play a pivotal role. Unlike traditional CPUs, GPUs are designed to handle parallel processing tasks more efficiently, making them ideal for AI workloads, such as training large language models and processing big data analytics.

In response to this need, data centers are adapting by incorporating more GPUs into their infrastructure, supported by innovations in rack design and cooling solutions. For instance, the use of liquid cooling and direct-to-chip cooling technologies is becoming more common, allowing data centers to maintain higher power densities without compromising on performance or safety​.

AI’s Impact on Data Center Operations

Artificial intelligence is not only driving the demand for more robust data-center infrastructure but is also transforming how these facilities are managed. AI-powered tools are being increasingly used to automate data center operations, offering capabilities such as predictive maintenance, resource optimization, and enhanced security measures. These AI-driven solutions enable data centers to operate more efficiently, reducing downtime and improving overall performance.

Moreover, AI helps in optimizing energy usage within data centers, a critical concern given the rising power consumption associated with higher density racks and the deployment of GPUs. By analyzing patterns and predicting future needs, AI systems can dynamically adjust cooling and power distribution, leading to more sustainable operations​.

Sustainability Challenges and Innovations

As data centers continue to expand, sustainability remains a pressing issue. The integration of GPUs and other high-density computing hardware significantly increases the energy consumption of these facilities. To mitigate the environmental impact, data centers are adopting advanced cooling techniques, such as closed-loop liquid cooling, which are more energy-efficient compared to traditional methods. Additionally, there is a growing focus on improving the energy efficiency of data centers through the use of renewable energy sources and more efficient power management systems.

These innovations are crucial as data centers are projected to account for an increasing share of global electricity consumption. Addressing these sustainability challenges is not only essential for reducing the carbon footprint of data centers but also for meeting the regulatory demands and corporate sustainability goals of their tenants​.

The Future Outlook: Sustained Growth and Emerging Risks

Looking ahead, the outlook for data-center landlords appears robust. As AI continues to evolve and become more deeply integrated into various industries, the demand for data centers is expected to remain strong. Even if AI adoption takes longer to fully materialize, the regular growth in data storage needs will continue to fuel demand. This scenario suggests that the era of weak rent growth is firmly behind us.

However, there are potential risks on the horizon. Investors are becoming increasingly concerned about the massive spending by tech companies on AI initiatives and are beginning to question when these investments will start to deliver substantial returns. A delay in the broader rollout of AI technologies could temper the rapid growth in demand for data centers. Nonetheless, given the tight market conditions and the steady increase in data storage needs, data-center landlords are well-positioned to weather any short-term fluctuations.

Partner with Vertical Data for Cutting-Edge Data Center Solutions

As the demand for AI and high-performance computing drives the evolution of data centers, ensuring your infrastructure is equipped to handle these challenges is crtiical. Vertical Data offers a comprehensive suite of solutions, from GPU procurement to high-density colocation services, designed to power next-generation data centers. With our expertise in sourcing hard-to-find equipment and optimizing infrastructure for AI workloads, Vertical Data is your strategic partner in staying competitive for the AI era.

Ready to elevate your data center capabilities? Partner with Vertical Data today to future-proof your infrastructure and stay ahead in the competitive compute market. Let's accelerate your growth together.