2026's Hidden Champions: 5 U.S. Stocks Quietly Powering the AI Era
GlobalFoundries, Lumentum, Corning, Analog Devices, and Keysight — the infrastructure plays the market hasn't fully priced in
In a market dominated by headlines about Nvidia and the mega-cap AI trade, some of the most compelling risk-adjusted opportunities in 2026 are sitting one layer deeper in the stack — in the materials, components, and test equipment that make AI infrastructure physically possible. Five names in particular have caught serious analyst attention: GlobalFoundries (GFS), Lumentum (LITE), Corning (GLW), Analog Devices (ADI), and Keysight Technologies (KEYS). Here's the data-driven case for each.
🔵 GlobalFoundries (GFS) — The Geopolitically Essential Foundry
Current price: ~$47.94 | Analyst avg. target: $50.62 | Rating: Buy
After years of being dismissed as a also-ran to TSMC, GlobalFoundries is having a genuine breakout moment. The company posted full-year 2025 revenue of $6.79 billion — up just 1% year-over-year on the surface, but with gross margin expanding 360 basis points to 29%. More importantly, the Q4 2025 earnings call revealed silicon photonics revenue effectively doubling again in 2026, targeting $1 billion from that segment alone by 2028. The company announced a $16 billion U.S. manufacturing investment and secured over 500 sole-source design wins worth more than $3 billion in combined lifetime revenue — specifically driven by customers who need GFS's three-continent footprint (U.S., Europe, Asia) for supply chain resilience. Automotive revenues surged 62% quarter-over-quarter. For Q1 2026, guidance is $1.625 billion in revenue with a 30% gross margin target. A $500 million share buyback has also been authorized. Valuation is split: the analyst consensus implies modest upside to ~$50, while a DCF-based view from Simply Wall St. puts fair value closer to $37. The near-term story is operating leverage + silicon photonics ramp; the longer-term story is geopolitical inevitability.
"Physical AI will outstrip the current data center boom." — GFS CEO Tim Breen, Q4 2025 Earnings Call
Fair Value: $48–$55 (12-month) | Growth Outlook: ★★★★☆ — Silicon photonics + AI infrastructure + geopolitical tailwind, three converging catalysts
🟣 Lumentum Holdings (LITE) — The Optical Rocket Ship
Current price: ~$563 | Analyst avg. target: ~$490–511 | Rating: Buy (13 analysts)
Lumentum has been one of the most stunning performers of the past 12 months — the stock has quadrupled from its 52-week low of $45.65 to above $560, and it's not without reason. The Cloud & Networking segment now represents 86% of total revenue, powered by surging demand for EML (Electroabsorption Modulated Laser) chips used in 200G and 400G data center interconnects. Q2 FY2026 (December quarter) delivered EPS of $1.67, beating consensus by 18%, with operating margin reaching 25.2%. For Q3, Lumentum guided EPS of $2.15–$2.35 versus Street estimates of $1.60 — a massive beat-and-raise. Revenue growth was 65.5% year-over-year. Analysts at Citi raised their target to $560, BofA to $600, and UBS to $455 following earnings. The bull case: hyperscalers are upgrading from 100G to 400G/800G infrastructure faster than anyone projected, and Lumentum is one of only a handful of companies that can supply the optical components at scale. The bear case: at current prices, the stock is already pricing in near-perfection, and any slowdown in hyperscaler capex would be punishing.
Fair Value: $490–$560 (analyst consensus) | Growth Outlook: ★★★★★ — Strong momentum, though current valuation already reflects significant optimism
🟡 Corning (GLW) — A 175-Year-Old Company's AI Windfall
Current price: ~$87–$133 (volatile post-Meta deal) | Analyst avg. target: $98–$122 | Rating: Strong Buy
Corning is one of the most counterintuitive AI plays in the market. Famous for making iPhone screen glass, the 175-year-old company is now a critical fiber-optic infrastructure supplier for AI data centers. The defining event of early 2026 was a deal with Meta Platforms for up to $6 billion in fiber-optic cable to power Meta's AI data center expansion — the stock surged 16% on the announcement, briefly hitting a new 52-week high above $110. Full-year 2025 showed revenue up 20.9% quarter-over-quarter with operating margin at 14.6%. The optical communications segment is the growth engine, while the display glass business provides a steady cash flow base. TipRanks analysts have a 3-month average target of $121.80 with a Strong Buy consensus (37 buy ratings, 15 holds, zero sells in the current period). The stock has returned 115% over the past 12 months — yet many institutional investors still classify it as an "old economy" materials company, creating a persistent valuation gap.
Fair Value: $98–$125 | Growth Outlook: ★★★★☆ — Structural fiber demand growth; the Meta deal is likely just the beginning
🟢 Analog Devices (ADI) — Cycle Recovery Meets the AI Edge
Current price: ~$322.97 | Analyst avg. target: $350.30 | Rating: Buy (20 analysts)
Analog Devices is the most "classic" of the five — a blue-chip semiconductor company that has been navigating an inventory correction cycle for the better part of two years. The good news: that cycle appears to be turning. ADI's core markets — industrial automation, automotive, communications infrastructure, and healthcare — are all seeing demand stabilization heading into 2026. The company's differentiation is its dominance in mixed-signal and data conversion chips, which are irreplaceable in the analog-to-digital interface layer of virtually every electronic system. Analyst EPS forecasts for FY2026 are $10.05 (up from $4.59 current), with FY2027 at $11.50, implying a recovery trajectory that puts the stock at roughly 32x forward earnings — reasonable for a business with this level of moat. The 20 analysts covering ADI are collectively bullish, with no sell ratings and a Buy consensus. ROE is forecast at 21.79%. The key risk is timing: if the industrial recovery stalls or takes longer than expected, the stock may underperform.
Fair Value: $330–$360 | Growth Outlook: ★★★☆☆ — Steady and reliable; a cycle recovery play rather than explosive AI growth
🔴 Keysight Technologies (KEYS) — The Referee of the AI Era
Current price: ~$236 (post-earnings rally) | Analyst avg. target: $222–$226 | Rating: Strong Buy
Keysight makes the electronic test and measurement equipment that engineers use to design and validate everything from 5G base stations to AI chip interconnects to quantum computing hardware. This is a "picks and shovels" play on technological complexity itself — and complexity is only increasing. Q4 FY2025 results were genuinely exceptional: revenue of $1.42 billion beat estimates, non-GAAP EPS of $1.91 topped forecasts, and orders grew 14% year-over-year to $1.533 billion — well ahead of the $1.43 billion the Street expected. The book-to-bill ratio of 1.08 signals strong near-term demand visibility. The Commercial Communications segment grew 12% on AI/data center demand; Aerospace/Defense and Industrial both grew 9%. For FY2026, analysts project EPS growth of 16.6% to $7.16. JPMorgan raised its price target to $255 on February 18, 2026, with an Overweight rating. Wells Fargo has a $225 target, Goldman Sachs $232, UBS $230. The secular growth drivers — 800G/1.6T silicon photonics testing, 6G research, and quantum computing validation — are all in very early innings.
Fair Value: $220–$255 | Growth Outlook: ★★★★★ — The more complex technology gets, the more indispensable this business becomes
Side-by-Side: How the Five Stack Up
Breaking the five names down by character: LITE and KEYS currently show the strongest direct AI momentum, but also carry the highest valuation risk — both are priced for near-flawless execution. GLW remains relatively underappreciated because the market still mentally files it under "old economy materials," yet the Meta fiber deal signals a structural re-rating is underway. GFS is the most direct beneficiary of geopolitical semiconductor supply chain diversification, with silicon photonics revenue growth as the key near-term trigger. ADI is the most defensive and stable of the group — a bet on an industrial cycle recovery rather than an explosive AI growth narrative. What unites all five is that they operate at the physical infrastructure layer of the AI build-out cycle, a layer that tends to be durable across multiple years and less susceptible to hype-driven reversals than pure software or chip design plays.
This article is for informational purposes only and does not constitute financial advice. Always conduct your own due diligence before making any investment decisions. Past performance is not indicative of future results.