Forget the Big Chipmakers for a Minute: This Overlooked AI Stock Is Up 114% and Just Landed Deals With Nvidia and Amazon

Corning’s optical communications sales rose 36% year over year in Q1 to about $1.85 billion, while core EPS increased 30% and operating margin reached 20.2%. The company also secured multibillion-dollar fiber agreements with Amazon and Nvidia, including a Nvidia partnership that could support up to $3.2 billion of investment and tenfold optical capacity expansion. Shares have more than doubled this year, but the stock now trades at about 90x earnings, tempering the otherwise strong AI-driven growth story.

Analysis

The market is starting to price Corning less like a mature materials supplier and more like a strategic toll road on AI cluster buildout. The second-order winner is not just GLW, but the entire optical stack: fiber preform capacity, connector vendors, and U.S.-based industrial automation/plant equipment suppliers that benefit as hyperscalers push supply-chain localization to reduce lead times and geopolitical risk. The flip side is that the more these orders concentrate into a handful of hyperscale customers, the more Corning’s growth becomes lumpy and headline-dependent rather than broad-based.

The key risk is that investors are extrapolating a multi-year demand curve from signed partnerships whose revenue recognition will likely stair-step over several quarters. That creates a classic mismatch: the stock can keep rerating on announced capacity, but fundamentals may not outpace expectations until the new plants actually ramp and yield improves. If AI capex pauses even modestly for one or two hyperscalers, the market will quickly re-rate the name because the current multiple leaves little margin for delayed conversions or pricing pressure.

The most interesting contrarian angle is that the “AI infrastructure” trade is shifting from semiconductors to bottlenecks with lower visibility and potentially lower moat durability. Corning’s density advantage is real, but if it proves modular and replicable, competitors can eventually attack on price once capacity expands across the industry. In that scenario, the right way to own the theme is not to chase the fully re-rated leader, but to express the view through under-owned enablers with earlier-cycle earnings leverage and less narrative premium.

Key Decisions for Investors

• Avoid initiating fresh outright longs in GLW here; if already long, trim 25-35% into strength and retain only a core position for a 6-12 month horizon. Risk/reward is asymmetric to the downside given the rich multiple and unclear revenue timing.
• Pair trade: long a basket of optical infrastructure suppliers / industrial capex enablers against GLW for the next 3-6 months to capture execution upside without paying the highest multiple in the group. Use this if you want AI infrastructure exposure but lower valuation risk.
• For tactical traders, buy GLW dips only on a post-earnings or post-capex headline pullback of 8-12% and use a tight 10% stop; the story likely remains intact over 12-24 months, but the near-term crowding is elevated.
• Own NVDA vs. GLW on a relative basis for 6-12 months if you want the cleaner monetization path; NVDA still captures the spend cycle earlier, while GLW is more exposed to delivery slippage and customer concentration.
• If pursuing upside in GLW, use call spreads rather than stock: 6-9 month bull call spreads limit downside if hyperscale spend normalizes while preserving participation if plant ramps and margin expansion continue.