AI Budgets Rebalance: Chips Are No Longer the Whole Story
June 16, 2026 — The AI boom has drawn attention to chipmakers, but the money trail is widening. In the latest market pulse, investors and corporate buyers are shifting emphasis from the headline-grabbing AI chips to the very hardware that keeps hyperscalers online: data-center infrastructure and related services.
Within this shift, the refrain forget chips. half every has echoed through investment rooms and conference calls. It’s not a rejection of AI semiconductors, but a recognition that a growing AI economy requires a robust backbone—power supply, cooling systems, and maintenance—areas that historically capture a larger share of datacenter spend than chip components.
Backlog and Margins Point to a Longer, Cheaper Path to Growth
One practical illustration of this dynamic: a leading data-center infrastructure provider disclosed a backlog near $15.0 billion, up about 109% from a year ago. The firm also reported a book-to-bill ratio close to 2.9x, signaling multi-year revenue visibility that stretches into 2027, independent of the latest GPU or CPU cycles. On the margin line, adjusted operating margin advanced by roughly 430 basis points to 20.8%, and adjusted earnings per share came in at $1.17, topping consensus expectations by about 16%.
Management used the momentum to raise full-year guidance for fiscal 2026, pointing to earnings per share in the $6.30 to $6.40 range as the company continues to scale capacity and diversify its revenue streams beyond hardware sales into services and lifecycle support. These numbers aren’t just quarterly beats; they reflect a steadier, asset-heavy growth path that can survive ups and downs in chip demand.
Why Infrastructure Wins in a Frosty Chip Cycle
Chip cycles can swing with supply, demand, and geopolitical tides. But the real cost of AI isn’t just the processor; it’s the energy, cooling, and resilience that allow AI workloads to run without interruption. Analysts point to several forces behind the shift:
- Power and cooling dominate datacenter spend, absorbing roughly four-fifths of total outlays in some segment analyses.
- Maintaining uptime and cooling efficiency becomes increasingly critical as AI models scale in size and complexity.
- Financing for capacity expansions is more favorable when a vendor carries an investment-grade rating and active index coverage, lowering the cost of capital for large-scale deployments.
With these levers in mind, investors are reimagining who benefits when AI budgets grow. Hardware manufacturers tied to power and climate control, data-center services firms, and suppliers of cooling and monitoring gear are gaining ground as durable growth stories in a market once dominated by chip firms.
Company Spotlight: Vertiv as a Pure-Play on the Backbone
Vertiv Inc. has emerged as a benchmark case for this broader reallocation. Positioned as a pure-play enabler of data-center resilience, the company’s backlog, margin trajectory, and revenue visibility look better aligned with a multi-year AI expansion than a quick chip cycle rebound. The company’s latest disclosures suggest that the majority of AI datacenter spending continues to stream into power distribution, cooling infrastructure, and services that optimize uptime and energy efficiency.
Industry observers highlight Vertiv’s backlog strength as a proxy for long-cycle capital expenditure. A more than doubling year over year confirms a steady stream of orders and project bookings that purchasers expect to convert into revenue across multiple years. In practice, this means robust demand visibility even if the appetite for cutting-edge AI chips moderates in the near term.
Capital Markets and the Financing Tilt
The capitalization environment around data-center investments has begun to reflect the longer planning horizons of hyperscale operators. Index additions and upgrades in оцен and credit ratings in recent months have reduced funding costs for capacity expansions and potential mergers and acquisitions in the sector. In stock market terms, these shifts help underpin a more durable growth profile for infrastructure players, regardless of the latest chip price movements.
Analysts who study AI supply chains say the market is moving away from a narrow chip focus toward a balanced view: chips still matter, but the most significant ROI comes from the efficiency, reliability, and scalability of the underlying datacenter ecosystem. As a result, a portion of investor capital that once chased chip leadership is reallocated to infrastructure and service-oriented firms that capture the long tail of AI deployment needs.
What This Means for Investors Right Now
For investors, the landscape is shifting in real time. Here are practical implications and watchpoints to guide portfolios through the coming quarters:
- Focus on backlog quality: A healthy order book with long-term service commitments signals revenue durability beyond quarterly GPU cycles.
- Assess margin resilience: Companies that can expand adjusted margins while maintaining service-based revenue streams may outperform during AI capex cycles.
- Evaluate financing dynamics: Credit ratings and index exposure that reduce capital costs can accelerate capacity expansion and strategic acquisitions.
- Monitor energy and efficiency trends: Innovations in cooling and power management remain pivotal to unit economics and sustainability goals.
In this environment, the investment thesis begins to look less like a single-chart bet on chip prices and more like a story about the entire AI datacenter ecosystem. And that is precisely the reason some market strategists are nudging clients to adopt a more nuanced allocation: not all AI dollars flow into chips; a meaningful share is earmarked for the infrastructure that makes AI work at scale.
The Bottom Line: A Shift That Could Reshape Winners and Losers
The AI revolution is still accelerating, but the way money gets deployed is evolving. While chipmakers remain relevant, the strongest, most persistent gains may come from players who secure the data-center backbone: power, cooling, and lifecycle services. The verdict from capital markets in the coming quarters will hinge on which firms can convert backlog into profitable, scalable growth while customers lock in cost-effective, energy-efficient capacity.
For readers and investors watching this transition, remember the rising adage forget chips. half every—a reminder that AI’s economic engine runs on more than silicon. If the AI era is truly about scale, the frontrunners will be those who master the full stack of data-center functionality, from hardware to the heat removed and the energy bills managed along the way.
Discussion