Tower Semiconductor sprinted higher in May after a quarter that answered a lot of investor questions: revenue and profit surprised to the upside, management rolled out hefty silicon photonics orders stretching into 2028, customers even pre-paid for capacity, and analysts rushed to lift targets. This article walks through the numbers, the contracts that matter, why silicon photonics is becoming central to AI hardware, and what the recent analyst optimism implies for the stock.
Shares of Tower Semiconductor jumped roughly 15.5% in May, a sharp move that reflected more than short-term momentum. The surge was tied directly to fresh operating results and forward-looking contracts that gave investors tangible proof of demand rather than vague promises. That kind of clarity tends to compress uncertainty and push valuation multiples higher, at least in the near term.
In the first quarter Tower posted revenue that rose about 15.6% year over year to roughly $414 million, while adjusted EPS climbed about 62.9% to $0.57. Both figures topped analyst expectations, and management followed up with a revenue guide of about $455 million for the current quarter, implying roughly 10% sequential growth. Those beats and the optimistic guide created a clear narrative that demand is real and the business is scaling.
What got the market especially excited was the disclosure of long-dated silicon photonics contracts totaling around $1.3 billion for 2027, plus customer prepayments earmarked to secure dedicated capacity. Tower also noted that its biggest customers have even larger commitments for 2028, with another round of pre-payments expected in early 2027. Customer-funded capacity removes a lot of execution risk for a foundry and gives management a runway to plan investments with more certainty.
Those prepayments and multi-year orders translate into medium-term revenue visibility that helps justify a higher valuation, though the stock now trades at a steep forward price-earnings multiple in the neighborhood of 70. A lofty multiple like that signals the market is paying up for growth expectations, not current margins, so execution against the 2027-2028 bookings will be the real test. If Tower hits the milestones investors anticipate, the high multiple could be rewarded; if it stumbles, the multiple could reverse quickly.
Susquehanna analyst Mehdi Hosseini responded by nearly doubling his price target from $180 to $330, citing management’s guidance and the pre-payment news as reasons to accelerate earnings forecasts. Hosseini projects an $8 EPS run rate by 2028 with upside to as much as $10 per share under favorable scenarios. Those estimates suggest a long runway for earnings expansion relative to the roughly $2.16 the company earned over the prior 12 months, which explains why some sell-side desks are growing bullish.
At the heart of the story is silicon photonics, an approach that uses light instead of copper to move data and is especially attractive as data rates and AI workloads explode. Copper links run into physical limits and increasingly rely on power-hungry digital signal processing to preserve signal quality at high speeds, which makes optical alternatives more compelling for large-scale AI farms. Even Nvidia has reportedly invested billions into companies tied to silicon photonics, underlining industry conviction that optical interconnects are central to next-generation compute infrastructure.
Tower occupies an interesting niche: it’s not a bleeding-edge logic foundry but it specializes in older nodes and crown-jewel specialty processes where silicon photonics can be manufactured reliably at scale. That positioning makes Tower a logical production partner for whoever consolidates the SiPho stack into high-volume deployments. Investors should weigh the upside from these contracts and investor enthusiasm against execution risk and the premium the market is already applying.
For anyone deciding whether to buy Tower now, the facts are clear: strong quarter, multi-year contracted demand with prepayments, aggressive analyst upgrades, and a strategic role in an area that matters to AI infrastructure. Those positives come with a valuation that assumes a lot of future success, so patience and a close eye on delivery milestones are sensible if you plan to own the stock. A measured approach recognizes both the opportunity in silicon photonics and the tightrope of high expectations ahead.
