This Nvidia-Backed Company Just Entered the $154 Billion AI Photonics Market. Does This Make the $13 Stock a No-Brainer Buy?

Goldman Sachs says AI-linked optical networking TAM could rise from about $15 billion in 2026 to $154 billion by 2028, with co-packaged optics representing a $91 billion opportunity. Nokia is expanding photonic-chip test and packaging operations and reported 49% growth in its AI & Cloud segment, 20% growth in Optical Networks, and 1 billion euros in new AI infrastructure orders in Q1 2026. The article also highlights roughly 108% year-to-date stock gains and insider buying, reinforcing a constructive view on Nokia's AI infrastructure exposure.

Analysis

This is less about a one-off Nokia rerating and more about the market repricing the bottleneck in AI capex from compute to interconnect. Once GPU clusters scale beyond a few racks, the value shifts toward firms that can remove watts per bit, shorten design cycles, and qualify packaging at hyperscaler tolerance levels; that tends to favor suppliers with both optical IP and manufacturing credibility, while commoditized module vendors face margin pressure as the ecosystem standardizes.

The second-order winner is probably not the obvious pure-play headline names, but the broader optical supply chain: packaging, test, advanced materials, and lithography-adjacent vendors that become gating items for volume deployment. If AI networking demand really compounds at the pace implied, lead times and qualification cycles should support operating leverage for the first movers, but also invite a wave of capacity adds that can compress returns by late 2027 if demand decelerates or alternative architectures reduce the need for discrete optics.

For Nokia specifically, the key question is durability of design wins versus narrative beta. The stock has likely pulled forward a meaningful amount of “AI networking optionality,” so upside now depends on conversion of orders into multi-year production ramps and evidence that optical margins improve rather than simply revenue mix. Insider buying helps sentiment, but it is not a substitute for proof that Nokia can defend pricing against larger semiconductor and networking incumbents once hyperscalers push for second-source redundancy.

The contrarian risk is that the market may be extrapolating CPO too aggressively before it is economically deployed at scale. Near-term, pluggables and incremental line-system upgrades can absorb more demand than expected, delaying the inflection for the most hyped architectures by 12-24 months; if that happens, the winners become the firms selling current-generation bandwidth rather than the ones priced for next-generation disruption.

Key Decisions for Investors

• Long GS vs short a basket of legacy telecom equipment names for 3-6 months: GS benefits from secular re-rating of the optical bottleneck, while slower-growth peers are more exposed to multiple compression if AI networking spend concentrates in best-in-class suppliers.
• Take profit tactically on NVDA upside exposure and rotate into optical infrastructure beneficiaries over the next 1-3 months: NVDA remains the capex anchor, but incremental upside may increasingly accrue to interconnect names as networking becomes the constraint rather than compute.
• Initiate a cautious long NOK with a defined stop, using 3-9 month horizon: the trade works if the market continues to capitalize AI networking optionality, but risk/reward deteriorates if new orders fail to convert into margin-accretive revenue.
• Prefer a pair long optical-enablement suppliers / short general telecom hardware over outright long NOK if available: this isolates the AI networking theme while reducing exposure to old-line telecom execution risk.
• Avoid chasing the stock after sharp momentum days; use pullbacks to build exposure, since the next catalyst is likely order conversion and margin evidence rather than additional narrative headlines.