Marvell Guided to 50% Growth and the Stock Dropped 6%. Every AI Investor Should Read That Warning.

Marvell Technology (NASDAQ:MRVL | MRVL Price Prediction) reported second-quarter fiscal 2027 results after Wednesday’s close, beat on revenue and earnings, guided the current quarter to roughly 50% year-over-year growth, and raised its outlook for both this fiscal year and next. Shares fell about 6% in after-hours trading.

CNBC’s Christina Parsons framed the disconnect on air: “The AI buildout is accelerating, not cooling off. The good CEO, Matt Murphy, raised the company’s revenue target for next year to roughly 18 billion.

That kind of growth landed against a stock up 184.54% year to date and 38.39% over the past month heading in. When a stock runs that far ahead of results, 50% growth can still disappoint. That is the mechanic behind Wednesday’s report.

Record Quarter, Raised Guidance, Falling Stock

Marvell posted Q2 revenue of $2.739 billion, up 36.5% year over year, with non-GAAP EPS of $0.94. Data center, at 79% of revenue, grew 46% year over year.

Q3 guidance came in at $3.15 billion plus or minus 5%, with non-GAAP EPS of $1.10 plus or minus $0.05. That is the roughly 50% growth number. Management raised the FY27 and FY28 revenue outlooks relative to prior guidance.

Matt Murphy tied the raise to demand. “AI-related bookings remain exceptionally robust, and we expect our revenue growth to accelerate further through the remainder of fiscal 2027,” he said in the release.

The stock traded down anyway. Shares closed the regular session at $241.45, then slid roughly 6% in extended trading. Operating results looked strong. Non-GAAP operating margin expanded to 36.6%, and net income rose 58.1% year over year. Disappointment sat on the demand side, as investors had already priced it in.

How Expectations Overtook Fundamentals

Growth, almost no company in the market can match, got sold because the price had already extrapolated further. That is the mechanic to internalize before the next AI earnings report lands.

Parsons said the same on air. “The sell off also a case, as we’ve seen with a lot of these names of high expectations, it was very owned or I should say it’s run up, what, 184% year to date into the print today, into the close, almost 30% just in the past month.”

Tool data confirms the direction, with MRVL up 47.77% between July 29 and August 27. Owned means price has room to disappoint.

Marvell’s history reinforces the pattern. Over the last seven reports, six were beats, yet the average day-of change was -2.09%. Clearing consensus and clearing an owned price are separate tests.

Alpha Vantage puts the trailing PE at 84x and the forward PE at 60x. At those multiples, expectations become the variable that determines returns.

How the Google Deal Got Priced In Twice

Marvell expanded a custom silicon partnership with Google earlier this year, including a warrant that allows Google to acquire up to 7% of Marvell’s shares contingent on revenue milestones. The stock reacted to that news on the day of the announcement.

By the time the revenue lands inside quarterly guidance, it is old information. Parsons flagged this directly. “The CEO said the revenue from that deal for next year is already baked into marvell’s numbers. It’s not exactly extra. The bigger payoff, he says, comes the year after, and that’s what cooled things off wall street.”

Buyers of the April announcement who also bought into this report paid twice for the same story. The structural problem is that meaningful economic contribution now sits a year further out. Owning a 60-plus forward multiple requires FY28 landing as guided and FY29 landing for the Google warrant economics to matter. That is a longer-duration bet on hyperscaler capex than most owners probably realize they are making (we profiled seven companies powering that same buildout, from optics to power to cooling, in a free AI infrastructure report).

What Marvell Sells, and Where the Real Risk Sits

Marvell sells optical interconnects, the parts that carry data between AI servers inside a cluster, and custom silicon, chips designed for a specific hyperscaler’s workload rather than a general-purpose merchant part.

Parsons summarized the mix. “The business is now growing more than 60%, driven by the optical parts that connect AI servers and by custom chips, which he expects to more than double.”

Optics became strategic because AI clusters outgrew the copper that once linked accelerators. When tens of thousands of GPUs sit in a single cluster, the connections between them become the bottleneck. Photonics relieves it, which is why Marvell bought Celestial AI and XConn earlier this year.

On custom silicon, Murphy said on last quarter’s call, “We remain confident in achieving our target model for our custom business to deliver on over $10 billion in revenue in fiscal 2029.” That number is what the FY29 story rests on.

The risk a 50% growth rate can hide is customer concentration. Data center is 79% of revenue, and the customers inside it are a handful of hyperscalers with a commercial incentive to vertically integrate over time.

Keep an eye on the stock into the October 6, 2026 Investor Day. If Marvell names FY28 custom programs with firm revenue attached, the sell-off reads as a valuation reset on a strong business. If hyperscaler capex commentary softens before then, concentration becomes the story and the multiple compresses quickly.

Contact [email protected] for any questions or corrections.