From GPU to Fiber: Why U.S. Optical Module Giants Are Surging
AI hardware shift from GPUs to optical modules as compute bottlenecks drive demand for high-speed interconnects, reshaping market valuations.
“AI Disruption” Publication 8900 Subscriptions 20% Discount Offer Link.
At the start of 2026, global capital markets remained transfixed by the AI wave. Yet while most investors were still debating whether AI computing demand had already priced in three years of future growth, and while headlines were dominated by NVIDIA’s latest chip launches, the optical communications sector in U.S. equities was quietly staging an independent and powerful rally.
Data showed that leading optical communications companies — Lumentum, Coherent, Corning — were hitting new all-time highs in succession, their performance clearly outpacing both semiconductor indices and top-tier AI software names. This phenomenon prompted deep reflection across the industry: within the grand AI hardware narrative, had a critical link been chronically undervalued by the market?
This was not a simple rotation of sentiment, nor merely a shift of capital from high-priced to low-priced assets. Rather, it was an industrial logic that had been underestimated for years, finally coming to fruition: against the backdrop of large-model scaling expanding exponentially, the true bottleneck had shifted from “computing chips” to “computing interconnects.” As data flows between GPUs grow geometrically, optical modules are no longer a supporting role — they have become the arteries and nervous system of AI infrastructure.
