There's an ETF for That? AI Data Centers

The Importance of Data Centers

Data centers are specialized facilities designed to house the computer hardware that powers cloud computing, artificial intelligence, and large-scale digital applications.

At their core, they consist of vast networks of high-performance servers, primarily outfitted with GPUs (graphics processing units), which are optimized for handling complex calculations simultaneously. These servers are stacked in racks, grouped into aisles, and maintained under strict temperature controls to prevent overheating.

Increasingly, AI companies are among the largest tenants, requiring enormous computing power to train and deploy machine learning models. AI’s hunger for infrastructure stems from the sheer scale of computation required. Unlike conventional software, AI models rely on processing vast amounts of data to improve their accuracy.

Training a state-of-the-art language model, for example, can take weeks or months and requires thousands of GPUs running in parallel. This is driving unprecedented demand for data center capacity. According to a McKinsey study, global demand for data center capacity could increase by 19% to 22% annually from 2023 to 2030.

To prevent a supply shortage, McKinsey says the industry would need to build twice the total data center capacity developed since 2000—in less than a quarter of the time.

A significant portion of this growth is being driven by cloud service providers. If you use Google Drive, Dropbox, or iCloud, you’re already relying on cloud storage. But the latest AI boom is merging AI capabilities with cloud infrastructure.

Companies like Alphabet, OpenAI, and Microsoft now use cloud-based data centers to power AI models such as Gemini, ChatGPT, and Copilot. These models require access to massive data sets for training, which must be stored and processed in secure, high-performance environments—making data centers indispensable.

The challenge now is keeping up. Legacy utility providers and data center operators are scrambling to meet surging power demands, source suitable locations, and upgrade infrastructure to handle the computational intensity. But obstacles remain—power shortages, rising latency concerns, and cooling requirements all present challenges to scaling up data centers at the pace AI development demands.

Pacer Data & Infrastructure Real Estate ETF (SRVR)

SRVR tracks the Solactive GPR Data & Infrastructure Real Estate Index, which is globally diversified and includes companies deriving at least 85% of their earnings from data center and infrastructure operations. Holdings are weighted by market capitalization.

Some of the ETF’s largest positions include Equinix Inc. (EQIX), American Tower Corp. (AMT), and Digital Realty Trust (DLR)—three U.S.-based real estate giants that own, operate, and manage extensive data center networks.

• Equinix specializes in colocation services, providing highly interconnected facilities where companies can store and process data.
• American Tower, originally a wireless infrastructure firm, has expanded into edge computing hubs to support AI-driven workloads.
• Digital Realty operates massive data campuses catering to cloud providers and enterprises with high-performance computing needs.

The ETF’s sector allocation reflects the interconnected nature of data infrastructure:

• 61% in real estate,
• 22% in communications,
• 13% in technology, and
• A small allocation to financials.

SRVR has an expense ratio of 0.55% and pays a 2.23% 30-day SEC yield on a quarterly basis.

Global X Data Center & Digital Infrastructure ETF (DTCR)

DTCR offers exposure to 25 holdings through the Solactive Data Center REITs & Digital Infrastructure Index. Like SRVR, its top positions include American Tower (AMT), Equinix (EQIX), and Digital Realty (DLR).

However, it also features Crown Castle Inc. (CCI), a company specializing in cell towers and small-cell networks, which play a crucial role in 5G deployment and edge computing—key enablers of AI-driven data transmission.

DTCR has a similar sector composition to SRVR, with 59% in real estate, but it tilts more toward technology with a 34% weighting.

This higher allocation to tech comes from key semiconductor and hardware companies involved in AI supply chains, including Super Micro Computer (SMCI), Intel (INTC), Lam Research (LRCX), TSMC (TSM), Nvidia (NVDA), Broadcom (AVGO), and AMD (AMD).

Compared to SRVR, DTCR is less of a pure-play REIT ETF and more diversified into the semiconductor and infrastructure side of AI development.

This lower focus on REITs and higher allocation to tech leads to a smaller 1.8% 30-day SEC yield, but DTCR has been a strong performer overall. It holds a five-star Morningstar rating and has delivered better risk-adjusted returns than most of the 211 real estate funds in its category.

This article is for informational purposes only and does not in any way constitute investment advice. The author may express their own opinions, which may not represent the opinions of ETF Central or its affiliated partners. It is essential that you seek advice from a registered financial professional prior to making any investment decisions.