Two AI Infrastructure Plays for the Next Growth Leg

The AI boom is entering a new phase. After years of focusing almost exclusively on compute power, the industry is now scaling around hardware-aware models due to a scarcity of chips. This shift is moving the investment focus from pure silicon to the next critical bottlenecks: memory and storage. The market is expanding rapidly to meet this demand, with four of the biggest tech companies projecting collective expenditures of $380 billion on data center and infrastructure build-outs this year.

While Nvidia's chips remain central, the steepest rallies in 2025 have come from other infrastructure plays. Stocks like Micron and Western Digital have more than tripled in value, outpacing even Nvidia's historic run. This reflects a market recognizing that fully operational AI data centers require a complex ecosystem beyond processors. They need memory, storage, fiber-optic cables, and central processors to function. As one expert noted, efficiency will be the new frontier, driving demand for the entire stack of hardware that supports AI workloads.

The bottom line for growth investors is that the TAM for AI infrastructure is not just about chips. It's about the entire plumbing that makes AI possible. As the industry moves from building compute to scaling it efficiently, companies providing the next frontier-memory, storage, and connectivity-are positioned for sustained growth. The focus has shifted from raw power to the reliable, scalable, and cost-effective solutions that keep AI systems running.

Micron (MU): The Memory Supercycle Play

Micron is the cornerstone of the AI memory supercycle, and its transformation from a cyclical laggard to a critical infrastructure play is complete. The company is a primary beneficiary of surging demand for high-bandwidth memory (HBM), a specialized DRAM variant essential for AI training. Analysts note we are very early in the memory cycle, with the total addressable market for HBM projected to hit $100 billion by 2028. This isn't just about incremental growth; it's a fundamental shift in the AI stack where memory is now the next frontier.

The financial impact is already visible. CNBC reports that DRAM prices surged in 2025 and are likely to increase further in 2026, with a predicted 40% rise this year. This price surge is directly fueling Micron's rally, which has seen the stock climb roughly 240% over the past year. Yet, despite this massive run-up, the valuation remains strikingly low. Micron trades at just 9.9 times forward earnings, a steep discount to the S&P 500's 22x multiple. This disconnect is what makes the PEG ratio so compelling: at a PEG of 0.6, the market appears to be pricing in the explosive growth trajectory, not the current earnings level.

The key to Micron's pricing power lies in the complexity of HBM production. Manufacturing these specialized chips consumes capacity that would otherwise be used for traditional, lower-margin products like smartphones and flash storage. This forces a difficult choice for competitors, giving Micron unprecedented leverage to secure fatter margins. As one analyst put it, "This is like getting a Mickey Mantle signed card at a garage sale" when it comes to the current price. For a growth investor, this setup offers a rare combination: a dominant position in a massive, secular TAM and a valuation that still reflects the early, supply-constrained phase of the cycle. The risk is that memory remains a commodity long-term, but for now, the bottleneck is real and Micron is at its heart.

Western Digital (WDC): The Storage Alternative

While memory prices are surging, Western Digital is positioned to benefit from the search for alternatives. As demand for high-bandwidth memory (HBM) drives DRAM prices up, AI companies may look to cheaper storage solutions to manage costs. This dynamic could be excellent news for Western Digital, which focuses on solid-state NAND and hard disk drives (HDDs) rather than DRAM. If the predicted 40% rise in DRAM prices this year materializes, the relative cost advantage of HDDs could make them a more attractive option for certain workloads, directly benefiting WDC's core business.

The company is already demonstrating strong growth, with its fiscal 2026 first quarter showing 27% year-over-year revenue growth. This expansion is fueled by its focus on the storage stack that supports AI, providing a scalable alternative to pure memory plays. For a growth investor, this setup offers a different kind of exposure: not to the memory supercycle itself, but to the infrastructure that must adapt as memory costs rise. Western Digital is building a business that scales with the overall AI build-out, but with a cost structure that could become more competitive under these new conditions.

The financial outlook supports this thesis. Analysts forecast 21% long-term earnings growth for the company, which, while not as explosive as some peers, is supported by a business model focused on durable storage demand. The key question is whether this growth can accelerate if memory bottlenecks force a shift toward hybrid or tiered storage architectures. For now, Western Digital provides a compelling counterpoint to pure memory plays, offering growth in a related but distinct segment of the AI infrastructure stack.

Catalysts, Risks, and What to Watch

The path to dominance in AI infrastructure will be determined by a few key variables over the coming year. For growth investors, the focus must shift from the broad theme to the specific supply-demand dynamics that will dictate which companies capture the most value.

First, monitor chipmaker memory production capacity and DRAM price trends. The recent 40% predicted rise in DRAM prices this year is a direct catalyst for memory suppliers like Micron. But it also creates a powerful counter-current for storage alternatives. As one expert notes, if DRAM prices rise, AI companies may seek cheaper alternatives, such as HDDs. This dynamic is the core of Western Digital's thesis. The company's ability to benefit hinges on this price pressure materializing and driving a tangible shift in procurement. Watch for quarterly updates from SK Hynix and Samsung, the other major DRAM producers, for signals on capacity utilization and pricing power. Any sign of supply glut could deflate the memory supercycle, while sustained tightness would validate the entire investment case.

Second, watch for announcements on AI factory expansions and dedicated power solutions. The industrialization of AI is moving beyond data centers to purpose-built facilities. Last year's $380 billion in projected infrastructure commitments from major tech firms is a starting point, but the next phase is about where and how that capital is deployed. Look for news on new AI campuses, like those rumored for Nvidia or Microsoft, and partnerships for specialized power grids. These projects signal a maturing, capital-intensive build-out where suppliers of the entire stack-chips, memory, storage, and connectivity-will be selected for long-term contracts. The companies that secure these anchor deals will likely see their growth trajectories accelerate.

The primary risk is a slowdown in hyperscaler capex. The entire infrastructure thesis rests on continued, aggressive spending from the major cloud providers. While enterprise AI spending is surging, as evidenced by $37 billion spent in 2025, the bulk of the infrastructure build-out is still driven by these few giants. Any hesitation from them, whether due to macroeconomic pressure or a recalibration of AI ROI, could deflate the market's appetite for new capacity. This would directly impact the revenue visibility for all infrastructure plays, from memory to storage. The market's recent rally, where stocks like Western Digital surged 15.8% in a single day on price news, shows how sensitive valuations are to these capex signals.

In short, the coming year will separate the durable growth stories from the speculative ones. The winners will be those whose business models are directly tied to the next wave of industrialization, not just the current memory price cycle. Keep a close eye on capacity, pricing, and the pace of factory announcements.