Two Undervalued AI Infrastructure Plays: A Value Investor's Perspective

The long-term structural tailwind for AI infrastructure spending is now a clear consensus. A report from asset manager Ninety One argues that the next wave of the tech revolution will rest on the hardware being built in Asia, with new agreements implying more than $US1 trillion in infrastructure spending before the decade is out. This is not a fleeting trend but a multi-year build-out that will require vast amounts of specialized components. For the value investor, the focus must be on identifying companies with durable competitive advantages-wide moats-that can capture value from this cycle, not just ride the short-term sentiment.

The key metric is the sheer scale of investment. A trillion-dollar infrastructure build-out creates a powerful demand-supply dynamic for critical components. This environment favors firms in non-commoditized segments where engineering expertise and manufacturing scale create pricing power. The noise in the market is often about the glamorous AI chips, but the real opportunity for long-term compounding may lie in the foundational hardware that enables them.

Two companies exemplify this thesis. Texas Instruments is positioned in the analog semiconductor space, a segment often overlooked but vital for AI systems. The company manufactures more than 80,000 crucial AI building blocks that are essential for power management, networking, and thermal control in data centers. Its deep integration into industries like automotive and industrial automation, where AI adoption is accelerating, provides a broad and durable demand base. While the stock has underperformed recently, its year-over-year revenue growth from data centers of more than 50% for the first nine months of 2025 signals a powerful inflection point.

Micron Technology is another prime candidate, operating in the high-bandwidth memory (HBM) market. This is a segment where demand is projected to outstrip supply for years, creating a classic imbalance. The company is well-positioned to benefit from this mismatch, as highlighted in a recent analysis. The value investor's lens here is not on quarterly memory price swings but on Micron's ability to maintain its technological edge and manufacturing capacity within a market that is structurally tight.

The bottom line is that the AI infrastructure opportunity is real, but it requires separating the signal from the noise. The trillion-dollar build-out is a long-term structural tailwind, but the winners will be those with wide moats in critical, non-commoditized segments. Companies like Texas Instruments and Micron are not pure-play AI stocks in the traditional sense, but they are deeply embedded in the physical foundation of the AI era. Their value proposition is one of durable demand and potential pricing power, which is the hallmark of a business capable of compounding over the next decade.

Texas Instruments: The Underappreciated Analog Engine

Texas Instruments operates a business model that is both foundational and non-disruptible. The company is the largest U.S. manufacturer of analog and embedded processing chips-components that are essential but often overlooked. These chips act as the vital "plumbing" for nearly every electronic system, from smartphones and vehicles to data centers and industrial equipment. This creates a wide and durable competitive moat. Unlike the high-profile, cyclical digital chipmakers, TI's analog business is characterized by long product lifecycles, high switching costs, and a deep trust from its customers. Its role as a dependable supplier for critical infrastructure makes its demand remarkably resilient across economic cycles.

The company's strategic investment underscores its commitment to this moat. Texas Instruments is planning a more than $60 billion investment across seven new U.S. semiconductor fabs, making it the largest investment in foundational semiconductor manufacturing in U.S. history. This massive build-out, spanning mega-sites in Texas and Utah, is designed to secure domestic supply chains for critical industries and position TI to meet the surging demand for its chips. While this capital-intensive expansion will pressure free cash flow in the near term, it is a long-term bet on the indispensable nature of analog and embedded processing.

A key growth signal is the explosive demand from the AI infrastructure sector. For the first nine months of 2025, Texas Instruments reported over 50% year-over-year revenue growth from data centers. This figure is so strong that management is now planning to report data center revenue as a separate segment, enhancing market visibility. This isn't a marginal trend; it's a fundamental shift where the company's core analog and power management products are becoming critical building blocks for AI systems, potentially turning it into a "virtual AI infrastructure pure play."

This operational strength is backed by a strong history of returning capital to shareholders. Despite the stock's recent underperformance, the company has raised its dividend for the 22nd consecutive year and repurchased $1.6 billion in stock in the first nine months of 2025. This consistent commitment signals deep confidence in the business's cash-generating ability and provides a tangible floor for the stock.

The investment case now hinges on valuation. The stock trades at a forward P/E of approximately 35.5, a premium that reflects its growth trajectory. This high multiple must be weighed against its growth and its role as a non-disruptible supplier. The evidence suggests the growth is real and structural, driven by AI, electric vehicles, and industrial automation. The $60 billion investment, while costly now, is a bet on securing that growth for decades. For the patient investor, the risk is not that TI will fail, but that the market will eventually recognize the full extent of its foundational role in the AI era. The current price may still be a reasonable entry point for a company with such a wide moat and a clear path to compound value.

Micron Technology: The Asymmetric Value Play

Micron Technology presents a classic asymmetric opportunity, where a powerful growth engine is priced for a much more modest story. The company's first-quarter fiscal 2026 results were nothing short of spectacular, with revenue soaring 57% year over year to $13.6 billion and earnings per share surging 167% year over year to $4.78. This explosive performance was driven by a fundamental demand-supply mismatch in the memory market, where AI and data center spending have far outpaced available supply.

The core of this growth is the high-bandwidth memory (HBM) market, a critical component for AI servers. Micron is well-positioned to capture this trend, with its 2026 HBM output already fully committed. The market itself is projected to expand dramatically, from $35 billion in 2025 to $100 billion by 2028. With a 21% share of this opportunity, Micron enjoys significant revenue visibility and pricing power, turning a cyclical commodity into a strategic, high-value product.

Yet the valuation tells a different story. Despite this growth, Micron trades at a forward price-to-earnings ratio of roughly 9, a stark contrast to peers like Broadcom, which commands a forward P/E of 35. This disconnect is the essence of the value play. The stock's recent 188% surge over the past 120 days highlights its volatility, but it also underscores how much ground the market has had to make up. The current price offers a tangible margin of safety relative to the projected growth trajectory.

The company is aggressively investing to meet this demand, having just lifted its fiscal 2026 capital expenditure plan to $20 billion. This capex is a double-edged sword: it funds the necessary capacity expansion to compound earnings, but it also increases capital intensity and future cycle risks. For a value investor, the key question is whether the current valuation adequately discounts these future risks while fully pricing in the near-term HBM tailwind. The evidence suggests it does not. Micron's combination of record growth, a fully booked near-term pipeline, and a valuation that still looks like a laggard's price creates a setup where the downside appears limited and the upside, from here, is substantial.

The Value Investor's Checklist: Assessing the Moat and Margin of Safety

For the value investor, the current market narrative around AI infrastructure stocks presents a classic dilemma: which companies offer a durable competitive advantage at a price that leaves room for error? The analysis points to two distinct profiles. Texas Instruments (TXN) and Micron Technology (MU) both stand to benefit from the AI boom, but their moats and valuations tell very different stories.

Texas Instruments possesses a wide moat built on essential, non-disruptible analog technology and immense manufacturing scale. Its chips-switches, sensors, buffers-are the vital, unseen components that keep data centers and automobiles operational. This is a moat of necessity, not novelty. The company's over 50% year-over-year revenue growth from data centers in the first nine months of 2025 is a powerful signal of this embedded demand. Yet, the stock has underperformed, trading at a forward P/E of roughly 35. This valuation embeds a premium for growth, leaving a narrow margin of safety. The company's $60 billion fabrication center expansion, while strategic, currently pressures cash flow. For the value investor, the question is whether the market will eventually recognize this as a pure-play AI infrastructure business, or if the growth premium is already fully priced.

Micron, by contrast, offers a moat defined by strategic capacity and pricing power in a constrained market. The company is a key beneficiary of the demand-supply mismatch for high-bandwidth memory (HBM), a critical AI component. With 21% share of a market expected to grow from $35 billion to $100 billion by 2028, Micron commands significant visibility and pricing leverage. Its recent operational execution is stellar, with revenue surging 56% year-over-year and net profit margins nearly doubling. The valuation here is the standout feature. Micron trades at a forward P/E of just 9, a fraction of peers like Broadcom (35x). This creates a substantial margin of safety, especially when paired with the company's record revenue growth and fully booked 2026 HBM capacity.

The owner earnings story further clarifies the picture. Texas Instruments demonstrates shareholder-friendly capital allocation, having raised its dividend for 22 consecutive years and repurchased $1.6 billion in stock. This supports a steady return of capital. Micron's capital allocation is more strategic and growth-oriented, with the company lifting its fiscal 2026 capital expenditure plan to $20 billion to capture the AI memory surge. This is not a passive return of cash, but an aggressive reinvestment into its moat.

The bottom line for the value investor is a choice between two paths. Texas Instruments offers a wide, essential moat at a premium price, requiring patience for the market to re-rate. Micron offers a powerful, strategically positioned moat at a bargain valuation, with a clear path to compound earnings. The margin of safety is far greater in the latter, making it the more compelling asymmetric opportunity for those willing to look beyond the glamour of the AI narrative.

Catalysts, Risks, and the Long-Term View

For both Texas Instruments and Micron Technology, the investment thesis is built on a powerful, long-term tailwind: the insatiable demand for semiconductors driven by artificial intelligence. Yet the paths to value realization are distinct, as are the risks that could derail them. A patient, multi-year perspective is essential.

Texas Instruments' primary catalyst is the execution of its historic $60 billion investment in U.S. manufacturing capacity. This isn't just expansion; it's a strategic bet to secure a dominant, low-cost supply of analog and embedded processing chips critical for AI infrastructure, automotive, and industrial systems. The company's deep partnerships with leaders like Apple, NVIDIA, and Ford provide a direct pipeline to this demand. A key near-term signal will be the ramp of its new 300mm fabs, which must translate into sustained revenue growth. Indeed, data center revenue has already shown over 50% year-over-year growth in the first nine months of 2025, a clear indicator of its AI integration. The long-term view is that this build-out will solidify TI's moat, turning its manufacturing scale into a durable competitive advantage.

The main risk for TI is the sheer scale and execution of this capital-intensive project. Funding a $60 billion build-out requires flawless capital allocation and disciplined spending, with free cash flow likely pressured in the near term. Any delay or cost overrun would directly threaten the return on this monumental investment. Geopolitical and macro risks are also present, including potential shifts in U.S. semiconductor policy and broader economic cycles that could dampen demand for the end markets it serves.

Micron's catalyst is the sustained, high-margin demand for memory, particularly high-bandwidth memory (HBM) used in AI servers. The company is well-positioned in a demand-supply mismatch, with its 2026 HBM output already committed. This creates exceptional pricing power and revenue visibility as the HBM market is projected to grow from $35 billion to $100 billion by 2028. The recent financial results underscore this: revenue surged 56% year-over-year in Q1 FY26, with net profit margins nearly doubling. The catalyst is clear: continued execution to meet this outsized demand while expanding margins.

The primary risk for Micron is cyclical memory oversupply, which could materialize if AI spending slows or if competitors rapidly increase capacity. The memory market is notoriously volatile, and a shift from scarcity to glut would crush margins and stock price. This risk is amplified by the company's own aggressive investment cycle. Macroeconomic downturns or geopolitical disruptions in key Asian manufacturing hubs could also impact supply chains and demand.

For both companies, the macro and geopolitical landscape is a constant undercurrent. Trade policies, export controls, and global economic health will influence both demand and the cost of capital for their massive expansion plans. The bottom line is that value realization for these AI enablers is not a quarterly event. It is a multi-year journey of capital deployment, technological execution, and navigating cyclical and geopolitical turbulence. The patient investor must look past near-term noise to assess whether these companies can successfully build and monetize their strategic advantages.