When the Nasdaq is flat and the equal-weight S&P is subtly outperforming the headline index by multiple points, there is a certain silence that has been hanging over the trading floors lately. You can hear it in the conversations as you pass any Midtown coffee shop in the morning: analysts hedging, portfolio managers shrugging, and retail traders checking the same five tickers. The trade in AI has not collapsed. It has simply grown more intricate. The majority of Wall Street hasn’t yet bothered to record the slower, stranger second story that is emerging beneath it.
It’s the convergence of AI and quantum. The actual plumbing of it—the lasers, switches, calibration software, and photonic components that only a few engineering teams consider—is what matters, not the version that is presented at conferences with holographic slides. There’s a feeling that those who construct the roads will be compensated long before those who drive on them. Usually, this is the case.
| Snapshot | Details |
|---|---|
| Theme | Quantum-AI Convergence — infrastructure buildout phase |
| Key Players Covered | Lumentum (LITE), Cisco (CSCO), Nvidia (NVDA), IonQ, Coherent |
| Market Size Forecast | $90B–$170B by 2040 (BCG); up to $1.3T economic value by 2035 (McKinsey) |
| Architectural Layers | Compute · Control · Networking |
| Notable Deal | Nvidia’s $2B strategic investment in Lumentum (March 2026) |
| Stock Movement | LITE up ~420% over the past six months |
| Risk Profile | Long runway, gradual revenue inflection, infrastructure-first |
| Investor Lens | Pick-and-shovel — own the suppliers, not the speculators |
| Time Horizon | Meaningful commercial advantage expected in the 2030s |
One of those technologies that seems uninteresting until you sit with it for a minute is Cisco’s universal quantum switch, which was unveiled earlier this fall. It maintains coherence, the delicate state necessary for quantum computing to function, while connecting superconducting and photonic quantum systems over regular fiber. Cisco, of all companies, was designed to solve this kind of unsexy interoperability issue. In the meantime, Nvidia has been subtly integrating AI into the control layer of quantum machines, employing machine learning to run error correction loops and stabilize qubits more quickly than humans could. Compute, manage, and connect. There are three layers, three distinct sets of suppliers, and none of them are yet priced like quantum stocks.
The cleanest example is lumentum. Because copper at hyperscale runs hot and wastes power, the San Jose photonics company uses light instead of copper to make the lasers and optical transceivers that transfer data through AI clusters. They signed a multiyear R&D partnership and received a $2 billion check from Nvidia in March. In just six months, the stock has increased by almost 420%.
It would be easy to label that as overheated, and perhaps it is. However, it is more difficult to refute the strategic reasoning. Lumentum is located at a chokepoint that is challenging to avoid, and hyperscalers are reorganizing their data centers around optical interconnects.
Investors appear to think that quantum networking will eventually follow the same physics. Photonic links are required for qubits to communicate across machines. Many of those components are produced by Coherent Corp., another player in the optics industry. Despite the commotion, IonQ is one of the few publicly traded companies that actually ships trapped-ion systems to business clients. These are all risky. Boston Consulting Group consistently reminds people that the revenue curve will be gradual. The $1.3 trillion estimate from McKinsey is for 2035, not 2027.
Even so, it’s difficult to ignore the similarities to the early days of cloud computing, when Equinix was just cooling and real estate and no one wanted it. Years ago, Tesla encountered similar skepticism. Seldom does the picks and shovels trade feel thrilling at the time. They are merely the trade that continues to be profitable after the gold rush is over. It’s still unclear if this one ends in 2030 or 2040. However, concrete pourers typically receive compensation regardless.
