Skip to content 
Everyone knows the Nvidia story. The chip giant, the trillion-dollar market cap, the GPU that powers the world’s data centers. It has become the face of the entire AI trade. But while investors were busy watching Nvidia, a different set of established, serious companies quietly delivered some of the biggest stock gains of the past year. Companies with real revenues, real customers, and real products. Companies that most people had simply never thought of as AI stocks.
Let us talk about them.
Lumentum: The Laser Company That Went Nuclear
Lumentum (LITE) has been around for decades. It makes lasers, optical components, and photonic chips. For most of its life, it was a steady, unremarkable business selling parts to telecom companies. Wall Street barely paid attention.
Then something changed.
As AI data centers began scaling to hundreds of thousands of GPUs, engineers ran into a fundamental physics problem. Copper wires, which connect servers inside a data center, cannot move data fast enough or efficiently enough at the speeds AI clusters demand. The answer is fiber optics. And fiber optics needs lasers. Lumentum makes those lasers.
The moment the market understood this, the stock took off. Lumentum is up over 1,269% in the past year alone, making it one of the best-performing stocks in the entire S&P 500. To put that in plain terms, a Rs 1 lakh investment made a year ago would be worth over Rs 13 lakh today.
The business numbers are just as striking. Revenue grew 90% year over year in its most recent quarter to a record $808 million. Operating margins expanded sharply. And then came the deal that confirmed everything: Nvidia announced a $2 billion investment in Lumentum to accelerate the development of next-generation optics for AI infrastructure. When the world’s dominant AI chip company writes you a $2 billion cheque, the market tends to pay attention.
Lumentum was recently added to the S&P 500, which means every passive index fund in the world is now required to own it. For a company that spent decades being ignored, that is quite the glow-up.
Vertiv: The Air Conditioning of AI
Here is a question most investors never think to ask. What happens when you pack thousands of AI chips into a room and run them at full power all day?
They generate enormous heat. A standard data center rack in 2023 consumed about 10 to 15 kilowatts of power. An AI-optimised rack today draws 120 to 150 kilowatts. That is a tenfold increase in heat density in roughly three years. Air cooling cannot handle that. The industry is being forced to switch to liquid cooling, and fast.
Vertiv (VRT) makes the power management systems, thermal management solutions, and cooling infrastructure that keep these data centers from melting down. It is, in the most literal sense, indispensable to the AI buildout. No cooling, no AI.
The stock is up 195% over the past year. Organic orders in its most recent quarter grew 252% year over year, which is a staggering number. Its backlog now stands at $15 billion, up 109% year over year. The company is guiding for $13.25 to $13.75 billion in 2026 revenue, representing roughly 34% growth. Analysts expect another 51% jump in adjusted earnings per share this year.
Vertiv was also added to the S&P 500 earlier this year alongside Lumentum, a clear signal that the infrastructure layer of the AI trade has gone from niche to mainstream. The data center cooling market is projected to grow nearly fivefold between 2025 and 2033.
Coherent: The Other Photonics Play You Missed
If Lumentum is the headline optical stock, Coherent (COHR) is its slightly quieter sibling that still managed to more than quadruple in price. The stock is up 367% over the past year, which in any other era would be considered a staggering return.
Coherent makes a broad range of optical and photonic components, including transceivers, lasers, and compound semiconductors. Like Lumentum, it sits at a critical chokepoint in AI infrastructure: the high-speed optical links that connect GPU clusters across data centers.
Nvidia made a significant investment in Coherent alongside its Lumentum deal, essentially signalling that the entire photonics supply chain needs to be strengthened and accelerated. Coherent reported quarterly revenue of $1.8 billion, up 21% year over year, with continued growth expected as demand for 800G transceivers accelerates.
The company was also added to the S&P 500 in the same March 2026 rebalancing that included Lumentum and Vertiv. All three additions in the same cycle told a clear story: the physical infrastructure of AI is no longer a side trade. It is the trade.
Arista Networks: The Invisible Backbone
Most people associate Arista Networks (ANET) with corporate networking. Boring stuff. Enterprise switches. The kind of hardware that keeps a bank’s internal servers talking to each other.
That reputation is outdated.
Arista has quietly become the dominant force in high-speed Ethernet switching for AI data centers. When you build a cluster with hundreds of thousands of Nvidia GPUs, those GPUs need to communicate with each other constantly and at enormous speed. The network connecting them is as important as the chips themselves. Arista builds that network.
The company reported $9 billion in fiscal 2025 revenue, up 29% year over year. Management has raised its 2026 AI networking revenue target from $2.75 billion to $3.25 billion. Its customers include Meta and Microsoft, who are building some of the largest AI clusters in the world.
What makes Arista particularly interesting is the competitive moat. It has now surpassed Cisco in data center switching market share, something that seemed nearly impossible five years ago. Networking accounts for roughly 10 to 15 percent of total AI cluster cost, but it is mission-critical, which means customers pay up for the best. Arista is, by most measures, the best.
The Pattern Behind All of This
If you step back, these four stocks share something important. None of them makes the AI model itself. None of them is building ChatGPT or Gemini. They are building the physical and network infrastructure that makes those models possible. They are the roads, the power grids, and the plumbing of the AI economy.
The investor insight here is straightforward. Every dollar that Nvidia, Microsoft, Google, and Meta spend on AI infrastructure flows through companies like these. The hyperscalers are investing hundreds of billions of dollars in data center buildouts, and a meaningful portion of that money lands in the balance sheets of Vertiv, Lumentum, Coherent, and Arista.
There is also a supply angle. AI is fundamentally a physical business. It requires chips, power, cooling, optical fiber, and network switches. All of these things take time to manufacture and deploy. Companies that are already at scale, already shipping to the major hyperscalers, and already embedded in the supply chain have a significant advantage. It is not easy to start making high-power lasers or liquid cooling systems from scratch.
The Caveat Worth Saying Out Loud
These stocks have already run hard. Lumentum trading at over 150 times earnings is not cheap. Vertiv’s valuation assumes the AI buildout continues at pace for several years. Any slowdown in hyperscaler capital expenditure, any change in AI infrastructure architecture, or any broader market correction could hurt all of these names.
This is not a story about finding hidden value. These stocks are well-discovered now. The question an investor must ask is whether the structural demand for AI infrastructure is durable enough to justify current valuations. Most analysts think the answer is yes, pointing to multi-year backlogs, rising order rates, and the sheer capital commitment being made by the world’s largest technology companies.
But the time when these were truly unknown stocks is behind us. What remains is the question of what comes next in the AI infrastructure buildout, and which company will be the next Lumentum sitting quietly in the background, waiting to be noticed.
That is always the harder question. And the more interesting one.
This article is for informational purposes only and does not constitute investment advice. Stock markets involve risk. Please consult a registered financial advisor before making investment decisions.
