Summary
Accelerating Compute: Why AI is Driving Data Center Demand
The promises of an AI-enabled future depend on deploying enough computational power, storage, and networking systems to train and run increasingly complex models. These models process enormous amounts of data, execute sophisticated algorithms, and make intelligent predictions or decisions by learning from this data. As AI expands into applications such as AI agents, autonomous vehicles, and humanoid robots, the need for high performance compute intensifies further, pushing demand well beyond current supply growth.
This demand imbalance is visible in the tremendous amount of money being directed at building data centers. Four of the largest hyperscalers—Alphabet, Amazon, Meta, and Microsoft—are together expected to spend nearly $600 billion to build data centers in 2026, up from approximately $350 billion in 2025 (Exhibit 1). With trillions more committed to future projects, the annual total global expenditure is forecasted to more than double by 2030.1 Meeting the expanding global demand for AI solutions will require a sustained, large-scale investment in AI infrastructure.
For investors, this level of spending represents an opportunity. While headlines about AI infrastructure often focus on who is doing the spending and the amount being spent, it is worthwhile to pay attention to where that money is going—who is being paid. Understanding where the money is going and the market dynamics driving that is incredibly relevant when deciding where to gain AI investing exposure.
EXHIBIT 1
Companies Are Committing Significant Capital To Build AI Systems
Source: Company data, Lazard Asset Management research, Bloomberg.
Notes: Amazon (AMZN) refers to AWS capex spend only, Microsoft (MSFT) data includes capital leases. Capex
spend estimates for 2025 and 2026 are Bloomberg consensus.
A Step Back: What Is AI Infrastructure?
To appreciate how and where value is being created and captured in AI Infrastructure, it is critical to take a step back and understand what goes into designing, constructing, and operating AI data centers. Although physically unremarkable from the outside, these facilities house complex, tightly integrated systems designed for high-density, low latency computing.
Data center infrastructure can be categorized into three main functions: compute, storage, and networking:
Compute
The hardware backbone that processes data and powers AI systems utilizing specialized chips (GPUs, TPUs), application-specific integrated circuits (ASICs), and highbandwidth memory (HBM)
Storage
The architecture dedicated specifically to data persistence and management
Networking
The interconnectivity fabric—comprising switches, routers, and cabling—that manages data flow within a data center and links to external end users
Surrounding this hardware is a broad collection of supporting systems and technologies that are critical to the efficient and continuous running of data centers.
For example, specialized software solutions play a key role at several levels—from chip design to cybersecurity to orchestration, monitoring, and optimization tools—to maximize performance and minimize disruptions. Similarly, industrial equipment is crucial to maintaining operating efficiency—data centers consume enormous amounts of electricity and generate significant heat, making power management and cooling solutions vital to keep servers running reliably.
These are just a few of the integrated layers of infrastructure that work in conjunction to deliver AI services and support applications. It is an extensive value chain of technology providers, with each piece serving a particular and necessary role in the process (Exhibit 2).
We believe a clear view of the value chain is critical when looking for investment opportunities. It allows us to chart where value is flowing and, more importantly, where it is accruing.
EXHIBIT 2
Understanding a company’s role in the AI value chain can help identify potential outperformers
What Matters: Understanding Supply and Demand Throughout the Value Chain
The mismatch between the unprecedented demand for AI and the time-consuming, capital-intensive expansion of infrastructure capacity is the result of several bottlenecks throughout the value chain:
Power
The exponential, 24/7 energy demand of new, high-density AI data centers is exceeding the ability of power grids to upgrade infrastructure. Data centers can be built in approximately 2 years, while the regulatory, permitting, and construction process for new power plants and high-voltage transmission lines could often take 5 to 10+ years, creating a massive bottleneck.
Memory
AI servers typically need 6x more dynamic random- access memory (DRAM) than standard servers do, driving up memory prices. Demand for HBM is accelerating even faster. Manufacturers are trying to improve supply either by improving efficiency or migrating to more advanced technology nodes, while expanding capacity is challenging given limited clean room space.
Leading-Edge Foundry
The ability to manufacture the most advanced semiconductors for AI is severely constrained. Constructing a new leading-edge semiconductor foundry (at 3 nm or 2 nm nodes) is costly (at around $20–25 billion per facility) and time-consuming and is further exacerbated by a shortage in specialized labor.
Limited availability and manufacturing capacity of these components and technologies are constraints in the global data center build-out.
However, many of the companies that operate on the supply side of these bottlenecks can benefit from greater pricing power, earnings leverage, and visibility into long-term demand.
Conclusion
While the AI narrative often focuses on the companies spending heavily on infrastructure, the more relevant question for investors should be where the value ultimately lands. As AI proliferation drives massive, multi-year investment across data center infrastructure, opportunities are emerging in the upstream layers of the value chain—particularly where supply is tight and replacement cycles are long.
Understanding how these dynamics have developed and are evolving will be key to identifying the businesses best-positioned to capture the next decade of AI-driven growth.
Notes
1. Dell’Oro Group, AI Boom Drives Data Center Capex to $1.7 Trillion by 2030, February 11, 2026
Important Information
Published on 1 May 2026.
The financial data presented is provided by external sources. Lazard Asset Management LLC takes reasonable care to ensure that the information provided is correct, but it neither warrants, represents, nor guarantees the content of the information nor does it accept responsibility for errors, inaccuracies, omissions, or inconsistencies.
Information and opinions presented have been obtained or derived from sources believed by Lazard Asset Management LLC or its affiliates (“Lazard”) to be reliable. Lazard makes no representation as to their accuracy or completeness. All opinions expressed herein are as of the published date and are subject to change.
Please consider a fund’s investment objectives, risks, charges, and expenses carefully before investing. For more complete information about Lazard ETFs and current performance, you may obtain a prospectus or summary prospectus by calling 800-823-6300 or going to www.lazardassetmanagement.com. Read the prospectus or summary prospectus carefully before you invest. The prospectus and summary prospectus contain investment objectives, risks, charges, expenses, and other information about the Portfolio and Lazard ETFs that may not be detailed in this document.
For mutual funds, each transaction involving a realized gain or loss is known as a taxable event. Taxable events thus accumulate throughout the year. At year-end, the fund’s investors may receive a taxable distribution or “capital gains distribution,” representing the gains generated by the fund’s transactions passed along to the investor.
This does not constitute legal, financial, or tax advice. We are not tax professionals. For specific advice applicable to your business, please contact a tax professional.
Equity securities will fluctuate in price; the value of your investment will thus fluctuate, and this may result in a loss. Securities in certain nondomestic countries may be less liquid, more volatile, and less subject to governmental supervision than in one’s home market. The values of these securities may be affected by changes in currency rates, application of a country’s specific tax laws, changes in government administration, and economic and monetary policy. Emerging markets securities carry special risks, such as less developed or less efficient trading markets, a lack of company information, and differing auditing and legal standards. The securities markets of emerging markets countries can be extremely volatile; performance can also be influenced by political, social, and economic factors affecting companies in emerging markets countries.
Certain information contained herein constitutes “forward-looking statements” which can be identified by the use of forward-looking terminology such as “may,” “will,” “should,” “expect,” “anticipate,” “target,” “intent,” “continue,” or “believe,” or the negatives thereof or other variations thereon or comparable terminology. Due to various risks and uncertainties, actual events may differ materially from those reflected or contemplated in such forward-looking statements.
Investors should be aware that these opportunities carry risks, including concentration in a narrow segment of the technology sector, sensitivity to shifts in AI capital spending, and potential regulatory or trade-policy changes affecting global supply chains.
This material is for informational purposes only. It is not intended to, and does not constitute financial advice, or an offer of financial products or to enter into any contract or investment agreement in respect of any product offered by Lazard Asset Management, and shall not be considered as an offer or solicitation with respect to any product, security, or service in any jurisdiction or in any circumstances in which such offer or solicitation is unlawful or unauthorized or otherwise restricted or prohibited.
The information provided herein should not be considered a recommendation or solicitation to purchase, retain, or sell any particular security. It should also not be assumed that any investment in these securities was or will be profitable.
Investment Products: NOT FDIC INSURED I NOT BANK GUARANTEED I MAY LOSE VALUE
The Lazard ETFs are distributed by Foreside Fund Services, LLC