When investors think about which companies form the bedrock of the artificial intelligence boom, they tend to reach immediately for the chipmakers — and for good reason. But the landscape of "foundational" AI plays has broadened well beyond a single name. The opportunity now stretches across several semiconductor companies and the software firms that make their hardware useful. Three chip names in particular stand out as core holdings: Nvidia, Qualcomm, and Intel, each playing a distinct role in the unfolding AI arms race.
Nvidia: Still the Backbone, Even Without Recent Momentum
Nvidia is the company most people associate with AI — the "king of AI." Yet lately its stock has been something of a head-scratcher for observers who have noticed a lack of upward momentum. Why hasn't there been much movement in the stock? Part of the answer is that the AI trade has simply broadened out to other companies, spreading attention and capital across a wider field of beneficiaries.
But that does not diminish Nvidia's central position. The company is no longer confined to GPUs; it is now moving into AI-driven CPUs as well, complementing its graphics processors. Nvidia will continue to drive the backbone of AI, and it is increasingly expanding into connected AI devices. The investment case is straightforward: Nvidia is the "must-own" stock in the bullish AI theme.
There is one structural caveat. Nvidia carries a tremendous market value, and at that scale it becomes mathematically harder to keep pushing the stock price higher — large numbers move more slowly. Even so, the stock should continue to climb over time and remains a core holding for anyone investing in AI.
Qualcomm: The On-Device, Edge-Compute Story
Qualcomm has seen a significant move up, driven by its push into manufacturing on-device AI processors. This points to where the next innings of the AI story are headed.
Is the next move in AI beyond the data center and cloud computing, toward on-device computing? Yes — the next frontier is on-device, connected compute at the edge. The vision is an evolution in which phones and connected devices have integrated AI functioning as an operating system. This AI would not merely answer questions; it would use built-in agents to act on information for the user, eliminating the need to open specific apps to make purchases or schedule things.
A concrete example illustrates the concept. Qualcomm is working with OpenAI, which last year hired Johnny Ive — the designer of the iPhone — to build AI-connected devices. They are reported to be releasing a phone next year featuring an AI-driven overlay operating system capable of answering questions. If a user wanted to schedule something or, say, purchase a movie ticket — finding out what time the movie plays — the system would go to the user's preferred app to complete the purchase, rather than forcing the user to navigate different apps manually. The result is an interactive, AI-driven process that lets devices carry out the activities the user wants completed.
Qualcomm is the pioneer of connected mobile devices and is integrating AI processors into more and more products. Reinforcing this trend, Microsoft recently launched a new line of Surface tablets that can be powered either by Qualcomm's Snapdragon AI processor or by processors from Intel, expanding into these AI-driven interaction capabilities.
What are the risks for Qualcomm? The principal risk is that the stock has struggled for a long time, despite the company's pioneering role and its growing integration of AI processors. In other words, being technologically early and well-positioned has not historically translated into reliable stock performance — there is a track record of the market underrewarding the company. That said, even with those past struggles, Qualcomm has been an outperformer so far this year.
Intel: Foundry Ambitions and Manufacturing Execution
Intel rounds out the trio and has been a dramatic outperformer recently — up more than 200% this year, and more than 400% on a year-over-year basis. It is also highly volatile: it fell about 8% in a single recent session before recovering some of that ground the following day.
Why remain bullish on Intel even after such a large run? The case rests on two levels:
1. Foundry / custom manufacturing. Intel is positioned to become the default manufacturer for companies like Apple and Nvidia, as well as other firms that want to make custom processors. This contract-manufacturing role represents a major new avenue of growth.
2. Core CPU business. Intel remains one of the largest producers of CPUs that go into data-center servers as well as desktop and laptop PCs — a durable, established franchise.
There is also fresh evidence of execution. Intel announced that it has entered risk production for the 18A variants of its manufacturing process — essentially the start of trial production. This marks steady progress toward delivering on its manufacturing roadmap, a key element of the foundry thesis.
The Experience That Will Surprise Us Most
Looking three to five years out in the AI arms race, which AI-powered experience will surprise people the most? It will be the interactive capabilities long depicted in television and movies — the ability to ask questions of an AI-powered device that interacts, engages, and then executes the processes the user wants. The reference point is the way Tony Stark talks to Jarvis: he engages, retrieves information, and Jarvis acts on his requests without requiring separate manual actions or separate applications. We will move toward an interactive process driven entirely by AI — "all AI all the time everywhere."
Software: The Indispensable Layer Behind the Hardware
Is there reason to worry about the software space — is it a part of the AI trade to dislike? No. The software space has come around, and it is essential to recognize that none of the hardware story works without software. None of it functions without the large-language-model database processors.
A leading example is Oracle, which continues to expand its ability to manage the huge databases that AI uses to gather and process information in order to deliver answers. Beyond Oracle, other attractive software names include MongoDB and monday.com. These stocks have struggled for a while but have recently begun to gain ground. The unifying theme is that it is software that provides the functionality to the hardware — the two are inseparable in delivering real AI capability.
The Through-Line
The foundational AI investment thesis is not a single bet but a layered one. Nvidia remains the indispensable backbone and a must-own core holding, even if its sheer size slows further price appreciation. Qualcomm represents the migration of AI from the cloud to on-device, edge computing — an emerging operating-system paradigm built around agents that act on the user's behalf. Intel offers a turnaround story rooted in becoming a default custom-chip manufacturer while retaining its dominant CPU franchise, with recent 18A manufacturing milestones validating its execution. And underpinning all of it is the software layer — Oracle, MongoDB, monday.com — without which the hardware delivers nothing. The destination is an interactive, agent-driven future where AI is woven into everything, everywhere, all the time.
