In his public talks about the future of artificial intelligence (AI), Qualcomm (QCOM +0.24%) CEO Cristiano Amon keeps coming back to one strong stance. Speaking at both Web Summit 2026 and Davos 2026, Amon argued that the AI race will be won at the edge, where devices, data, and users converge. Moreover, he says, success in this space will depend on constant reinvention, efficient on-device computing, and competition across both edge and data center inference in a market that may be overhyped in the short term but could grow to be far larger than expected over the long term.
Amon’s claims are bold. But the more I look at what Qualcomm is building, the more I think he’s right, and the more I think the market is missing it.
Despite negative news that has pushed Qualcomm stock down 25% year to date as of Thursday, the company last month signaled confidence in its future by announcing that it had authorized a new $20 billion share buyback and a quarterly dividend increase from $0.89 per share to $0.92 per share, further strengthening its position as a leading income stock in the tech space.
Image source: Getty Images.
What ‘edge AI’ means
Most of the AI conversation in 2026 has centered on data centers — the hardware that goes into them, such as Nvidia’s graphics processing units (GPUs), and the heavy computing workloads they handle in cloud inference and model training.Â
But there’s a growing argument that the real unlock of value for AI will happen when it runs on local individual devices and end-user systems — your smartphone, your car, a machine on the factory floor. That’s edge AI. When an AI model can process your request on your laptop instead of routing it to a server in Virginia, you get lower latency, better privacy, and no cloud bill. As Amon puts it, the user interface is becoming human-centric. Your device learns your context, your habits, your intent, in real time, without a round trip to the cloud.
This shift is not theoretical. At CES 2026, Qualcomm debuted the Snapdragon X2 Elite and X2 Plus processors for PCs. These chips are purpose-built to run AI agents locally, with 85 TOPS (tera operations per second) of on-device AI computing power. Industry analysts now project that laptops built around Arm Holdings’ CPUs, the category where Snapdragon X dominates, could capture 20% to 25% of the total market of machines running the Microsoft Windows operating system by the end of 2027. That’s a meaningful share of a massive market, and Qualcomm sits at its center.
Today’s Change
(0.24%) $0.31
Current Price
$128.06
Key Data Points
Market Cap
$137B
Day’s Range
$127.85 – $130.61
52wk Range
$121.99 – $205.95
Volume
8.4M
Avg Vol
11M
Gross Margin
55.10%
Dividend Yield
2.78%
The Dragonwing bet is the real story
When most people think of Qualcomm, they think of its Snapdragon chips and smartphones. But the company is building what may be one of the most underappreciated platform businesses in tech right now. At CES 2026, Qualcomm also unveiled the Dragonwing IQ10 Series. This is a processor explicitly designed as the “brain of the robot,” intended to power everything from industrial autonomous mobile robots to humanoid robots.
In March, Qualcomm signed a long-term collaboration with Neura Robotics to jointly develop what they called “brain + nervous system” reference architectures for next-generation humanoid robots. Such devices could help make AI a physical presence in people’s everyday lives.
Qualcomm’s automotive business tells a similar story. In its fiscal 2026 first quarter (which ended Dec. 28), the company’s automotive revenue hit $1.1 billion, up 15% year over year. It was that metric’s second consecutive quarter above $1 billion. The automotive design-win pipeline (contracts already signed but not yet generating revenue) now stands at $45 billion, representing future revenue from many major industry players.
The importance of design wins is something that investors need to recognize: These deals are commitments from automakers to use Qualcomm silicon in vehicles that won’t reach dealers for two or three years. By its fiscal 2029, Qualcomm is targeting $22 billion in combined automotive and Internet of Things (IoT) revenue.
This is where it gets interesting to me as an investor. As of early April, Qualcomm trades at around $125, roughly 22% below the consensus analyst price target. Its forward price-to-earnings ratio (P/E) is around 12, below that of most AI-focused chip stocks. Qualcomm’s record Q1 fiscal 2026 revenues in automotive and IoT came in alongside 24% year-over-year earnings per share (EPS) growth. That’s a company executing well — and still trading at a discount.
Here’s a risk to consider
The risk is real: Qualcomm’s biggest source of revenues is still the smartphone market, which is maturing. Apple is working on its own modems. Competition in the PC CPU space from Intel and AMD is intensifying. If the speculative timelines for humanoid robot adoption or automotive AI rollouts prove to have been too optimistic, that would put pressure on those ambitious pipeline figures.
But I think the market is still pricing Qualcomm as a phone chip company with AI exposure, when it’s increasingly becoming an edge AI infrastructure play across PCs, robots, cars, and industrial devices. I’m anticipating that eventually, the market will reprice it accordingly, which is why I’m buying the stock.