The 2026 semiconductor rally has been historic, and lopsided. Memory names like Micron have ripped more than 80% in a month. AMD is up roughly 110% year-to-date. And Intel is up nearly 500% in the past year. But under the surface, breadth is thin: many chip stocks are flat or down, and the sector’s biggest dividend payers have largely missed the bid. With IDC forecasting global semiconductor revenue near $1.3 trillion in 2026, fueled by AI infrastructure spending, the question for income investors isn’t whether the rally is real. It’s which laggards offer a credible path to participate.

Below are three chip-economy dividend stocks that have under-performed the headline AI complex but have visible 2026-2027 catalysts that could close the gap. Each pays a real dividend, has a multi-decade payout track record or strong balance sheet, and has exposure to a piece of the AI build-out that the market hasn’t fully priced.

1. Qualcomm (QCOM): The $20 Billion Buyback Hiding in Plain Sight

Why it’s lagged
Qualcomm has been the analog (no pun intended) for “good business, wrong story.” The market spent 2025 punishing names tied to handset cycles, and Qualcomm derived the bulk of its revenue from smartphone modems and SoCs. Memory constraints and a sluggish premium-Android cycle pushed handset revenue down 13% year-over-year in the company’s most recent quarter. Up 28% YTD sounds fine until you compare it to AMD’s roughly 110% gain or Micron’s surge — Qualcomm’s multiple has actually compressed, not expanded.

Why it could catch up
Three things have changed in the last 60 days. First, Q2 FY2026 results showed automotive revenue hit a record $1.33 billion, up 38% year-over-year — a high-margin, design-win-driven business that the sell-side still treats as a rounding error. Second, CEO Cristiano Amon said on the call that a “leading hyperscaler custom silicon engagement” is on track to begin shipping later this calendar year, signaling Qualcomm is finally getting a seat at the data-center table that has been Nvidia’s alone. Third, multiple reports place Qualcomm and MediaTek as co-designers of a custom processor for OpenAI’s rumored AI-agent phone, with Ming-Chi Kuo flagging a 2027 launch and shipment targets in the hundreds of millions.

On May 8, Daiwa upgraded Qualcomm to Outperform and lifted its price target to $225 from $140. Management also authorized a fresh $20 billion buyback on top of the dividend. For a stock at roughly 22x earnings — about half Nvidia’s multiple — the setup is asymmetric: you’re paid to wait while three independent catalysts (auto, hyperscaler, AI phone) try to re-rate the multiple.

The risk
The Apple modem business unwinds in 2027. If hyperscaler shipments slip or the OpenAI partnership doesn’t materialize on the rumored timeline, Qualcomm is back to being a smartphone story.

2. Skyworks Solutions (SWKS): Cheap, Hated, and Pivoting

Why it’s lagged
Skyworks is the cleanest “everything that hurt smartphone suppliers in 2025” story in the sector. RF front-end exposure, Apple concentration, and a brutal post-COVID phone cycle pushed the stock to roughly 28% off its 52-week high. The dividend yield, near 3.9%, is the highest among any U.S.-listed pure-play chip name of meaningful size. That yield exists because the market doesn’t believe it.

Why it could catch up
Q2 FY2026 results actually beat: $944 million in revenue versus $902 million expected, with EPS of $1.15 versus $1.04. More importantly, management disclosed a strategic design win with a “leading Android OEM” expected to generate over $1 billion in revenue through 2030. That’s the first crack in the iPhone-dependence narrative in years. The stock popped roughly 13% on the print and is, despite that, still well off its highs.

The longer thesis is edge AI. As on-device inference moves from cloud to phone — exactly the architecture OpenAI and Qualcomm appear to be designing for — RF complexity per device increases. Skyworks owns share in that complexity. If the AI-phone narrative turns into shipments in 2027, Skyworks gets a second leg of multiple expansion that’s not yet in numbers.

The risk
This is the messiest name on the list. The trailing payout ratio is reportedly above 100%, meaning the dividend is being funded partly from the balance sheet during the trough. Investors should size accordingly and watch the next two earnings reports for confirmation that free cash flow is recovering. A dividend cut isn’t the base case, but it’s not zero either.

3. Microchip Technology (MCHP): A Cyclical Comeback With a 22-Year Streak

Why it’s lagged
Microchip spent 2024 and 2025 absorbing one of the worst inventory corrections the analog/MCU industry has ever seen. Channel destocking gutted revenue, gross margins compressed, and the stock spent much of last year being treated as a value trap. Even now, the dividend yield near 1.8% is below the analog peer group, which keeps it off most income screens — and is exactly why it’s interesting.

Why it could catch up
The cycle is turning. Industry analysts are calling Microchip a “revitalized contender” entering Q2 2026, with management leaning into Silicon Carbide power devices, Edge AI MCUs, and a “Total System Solutions” go-to-market that bundles connectivity, security, and analog. The board just declared its 93rd consecutive quarterly dividend, the 22nd straight year of increases — the kind of capital-return discipline that gets rewarded when earnings re-accelerate.

The AI angle here is unglamorous but real: industrial automation, robotics, automotive ADAS, and edge inference all need the kind of mid-tier MCUs and analog interface chips Microchip sells. Every Nvidia GPU sold into a factory or autonomous vehicle drags along dollars of Microchip content. As CapEx broadens out from hyperscalers to industrial AI in 2026 and 2027, that content per system goes up, not down.

The risk
Cyclicals are cyclical. If the analog recovery stalls or the company over-earns into the next downturn, the yield gets re-rated wider — meaning a lower stock price. Watch inventory days and book-to-bill in the next two quarters.

The Bottom Line
The 2026 chip rally has been narrow enough that “left behind” is a long list. The three names above share a common shape: real dividends, real AI exposure, and a market that has so far refused to give them credit for the second part. Qualcomm offers the cleanest catalyst stack and a fortress balance sheet. Skyworks offers the highest yield and the most contrarian setup. Microchip offers the longest payout history and a textbook cyclical recovery.

None of these will go up like Nvidia did from 2023 to 2025. But for income-oriented investors who want chip exposure without paying 40x earnings for it, the laggards are where the math works.

Own the real estate AI is built on

Everyone wants to own the next Nvidia. But here’s a different approach: own the landlord.

Every AI company needs power, cooling, and connectivity. They’re spending billions to build out infrastructure – and someone has to provide it.

Robert Rapier invests in the companies that collect rent from the AI boom, not the ones betting everything on the next breakthrough.

See his current portfolio here.