Marvell Q2 Earnings Analysis: Why MRVL Stock Fell Despite Record Revenue

Marvell delivered almost everything investors normally want from an earnings report: record revenue, accelerating data-center growth, better-than-expected Q3 guidance and higher long-term forecasts. 

Yet MRVL stock fell 7.8% in after-hours trading. The problem was not weak execution. It was that a very good quarter still fell short of the enormous expectations created by Marvell’s Google agreement and the stock’s nearly threefold rise in 2026.

Let’s break down what Marvell reported, why investors were still disappointed, what the Google agreement adds to the story, and the numbers that should matter from here. The short version is simple: Marvell’s business passed the test, but its earnings surprise did not.

Marvell Q2 FY2027 Earnings Results: Revenue, EPS and Data-Center Growth

Marvell reported fiscal second-quarter revenue of $2.739 billion, 37% higher than a year earlier and 13% above the previous quarter. Revenue was $39 million above the midpoint of management’s guidance.

Adjusted earnings reached $0.94 per share, one cent above the FactSet consensus of $0.93. Data-center revenue grew 46% to a record $2.17 billion and represented 79% of total revenue.

According to Marvell’s official earnings release, operating cash flow was $605.5 million, while adjusted operating margin expanded to 36.6%.

The headline numbers were positive, but the size of the beat was small. Revenue exceeded the $2.72 billion FactSet estimate by roughly $19 million, or just 0.7%. The EPS beat was approximately 1.1%.

That is closer to clearing a hurdle with one shoe brushing the bar than sailing over it.

Marvell Q2 Results vs Our Pre-Earnings Benchmarks

In our Marvell earnings preview, we argued that a headline beat might not be enough. The report needed to pass three larger tests:

1. Data-center revenue needed to approach $2.15 billion.
2. Q3 revenue guidance needed to reach at least $3 billion.
3. Operating margin needed to improve towards 36.5%.

Here is how Marvell performed:

The preview’s operating model was remarkably close. Our estimated data-center revenue was only about $30 million below the reported figure, a difference of 1.4%. The estimated operating margin was within 0.1 percentage point of the actual result.

Marvell therefore passed the operational tests. The missing piece was the one investors had added after the Google announcement: evidence that the agreement created substantial new revenue before fiscal 2029.

Why Is Marvell Stock Falling After Earnings?

MRVL closed August 27 at $241.45, down 1.49% during regular trading, before falling another 7.8% after hours, according to Reuters and MarketWatch.

Four factors explain why a strong report produced a negative reaction.

1. The beat was too small for the expectations

Marvell beat revenue estimates by only 0.7% and EPS estimates by approximately 1.1%.

For an ordinary semiconductor company, that might be enough. But Marvell stock had gained nearly 170% during 2026 before the report as per Google Finance. Investors were paying for accelerating AI growth, not simply an in-line quarter.

The distinction is between the published consensus and the market’s informal “whisper number.” Published estimates said $2.72 billion. The stock’s run suggested investors were hoping for something meaningfully higher.

2. Much of the Google opportunity was already included

This was the biggest disappointment.

Marvell recently expanded its commercial relationship with Google across custom AI accelerators, networking, memory and storage products. The associated warrant gives Google the right to purchase up to 58.97 million Marvell shares at $206.58 each.

However, management clarified on the earnings call that revenue from covered programs through fiscal 2028 was already included in Marvell’s previously provided customer-revenue target.

In other words, the agreement increased confidence in the existing forecast, but it did not suddenly place a large new block of revenue on top of the FY2028 outlook.

That is the difference between validation and addition. Google validated Marvell’s custom-silicon capabilities. Investors, however, were hoping for substantial addition to near-term forecasts.

3. Investors wanted FY2029 numbers

Management said the Google agreement should become more significant in fiscal 2029 and beyond. Marvell also expects its custom-chip business to accelerate sharply during that period.

But CEO Matt Murphy stopped short of providing an updated FY2029 target. The company plans to share more at its October 6 Investor Day.

The Wall Street Journal noted that the absence of FY2029 guidance was one reason the report failed to deliver the surprise investors wanted.

The market was essentially told, “The larger opportunity is coming, but wait until October for the numbers.” After such a large stock-price run, waiting was not the answer some investors wanted.

4. Gross margin is expected to decline

Marvell’s adjusted gross margin was 58.9% in Q2. Management expects a Q3 range of 57.5% to 58.5%, producing a midpoint of 58%.

That is a sequential decline of 90 basis points.

CFO Dan Durn explained that the expected acceleration in custom chips would create a gross-margin headwind. Custom-chip programs can generate enormous revenue, but their percentage margins may be lower than those of certain connectivity products.

Think of it like a restaurant adding a large corporate catering contract. Revenue and total profit can rise, even if the profit earned on each dollar of catering sales is lower than the margin on individual meals.

Marvell Q3 FY2027 Guidance: Revenue, EPS and AI Growth Outlook

The Q3 guidance was stronger than the after-hours reaction suggests.

Marvell expects revenue of $3.15 billion, plus or minus 5%, compared with the company’s previous indication of approximately $3 billion. At the midpoint, revenue would increase 15% sequentially and more than 50% year on year.

Adjusted EPS is expected to reach $1.10, plus or minus $0.05.

The FY2027 outlook increased by 4.3%, while the FY2028 forecast increased by 9.1%. These are meaningful upgrades, especially when applied to multibillion-dollar revenue bases.

Management now expects total revenue to grow around 45% in FY2027 and another 50% in FY2028. The custom business is expected to more than double during FY2028.

One hidden number is particularly important. Management expects Q3 data-center revenue to grow by more than 20% sequentially. Applying 20% growth to Q2’s $2.17 billion produces at least $2.60 billion in Q3 data-center revenue.

That would mean data centers contribute more than 82% of Marvell’s Q3 revenue. Marvell is becoming a more focused AI-infrastructure company, but it is also becoming more dependent on one end market and a limited number of large customers.

Marvell Earnings Call: Key Comments From CEO Matt Murphy

Murphy said AI-related bookings remained “exceptionally robust” and that Marvell expects revenue growth to accelerate through the remainder of FY2027. The drivers are broader than custom AI accelerators:

• Demand for 800G optical products remains strong.
• Marvell’s 1.6T optical business is ramping rapidly.
• Scale-out switching revenue remains on track to more than double.
• Demand for transimpedance amplifiers and optical drivers is exceeding expectations.
• Custom XPU and XPU-attached products should accelerate in the second half.
• Scale-up optical and switching opportunities are expanding as AI clusters become larger.

This matters because Marvell is not betting on a single chip. Its portfolio includes the processors, switches, optical components, memory interfaces and network controllers that help thousands of AI processors work together. If AI accelerators are the engines, Marvell sells an increasing portion of the roads, traffic lights and loading docks around them.

Murphy also said the Google agreement, together with other programs, increased management’s confidence in the custom business for FY2029 and beyond. The October Investor Day will now carry unusual importance because that is where investors expect the confidence to be translated into numbers.

Marvell Q2 Earnings: What Went Right and What Disappointed Investors

Operationally, very little went wrong during Q2. The better question is: what failed to improve as much as investors expected?

The quarter did not break the business thesis. It weakened the surprise thesis.

That distinction is important. A falling share price does not automatically mean the company reported deteriorating fundamentals. It can also mean the market had priced in an even better outcome.

Marvell Gross Margin vs Operating Margin: What the Q3 Math Shows

A lower gross margin does not automatically mean Marvell will earn less operating profit. Using the midpoint of Q3 guidance:

Despite the expected gross-margin decline, adjusted operating margin could increase from 36.6% to approximately 37.2%.

Why? Revenue is growing faster than operating expenses. This is operating leverage.

Marvell raised its FY2027 adjusted operating-expense outlook from $2.45 billion to $2.55 billion, an increase of $100 million. Over the same period, its revenue outlook increased by $500 million.

That works out to roughly $0.20 of additional planned operating expense for every $1 of additional forecast revenue. If Marvell maintains a gross margin near 58%, a meaningful portion of the incremental revenue can still reach operating profit.

Management expects adjusted operating margin to enter its 38% to 40% long-term range during Q4 and move towards the upper end during FY2028. That is now one of the most important promises to verify.

Marvell Cash Flow, Stock Compensation and Share Dilution

Marvell generated $605.5 million in operating cash flow. Subtracting $126.7 million of property and equipment purchases produces a simple free-cash-flow proxy of approximately $479 million.

That equals roughly 17.5% of revenue.

Cash flow was below adjusted net income of $865.9 million, partly because Marvell is making supplier-capacity prepayments. Management expects approximately $1 billion of such payments during FY2027 to secure the manufacturing supply required for future growth.

There is also a gap between adjusted and GAAP earnings. Stock-based compensation reached $326.2 million in Q2, compared with $153.6 million a year earlier. Diluted weighted-average shares increased from 870.4 million to 921.2 million, or approximately 5.8%.

Adjusted net income grew about 48%, but adjusted EPS grew 40%. The rising share count absorbed part of the profit growth on a per-share basis.

Marvell repurchased $200 million of stock during the quarter, but investors should continue monitoring whether repurchases offset the dilution created by employee equity, acquisitions, preferred shares and potentially the Google warrant.

Marvell’s Google Warrant: What the $120 Billion Threshold Means

The warrant agreement filed with the SEC allows Google to purchase up to 58.97 million Marvell shares at $206.58 each.

Only 1.36 million shares vest mainly with time. The other 57.61 million are performance-based, with a portion vesting whenever Marvell records another $500 million of qualifying revenue. The agreement contains 240 revenue-linked tranches.

Multiplying 240 by $500 million produces a theoretical cumulative revenue ceiling of $120 billion.

But this is not a $120 billion order, backlog figure or revenue forecast. It is simply the amount of qualifying revenue required for all performance-based warrants to vest.

A useful way to read the agreement is through three layers:

The Google deal remains strategically significant. What changed after earnings is the timing investors should use when modelling it.

Should Potential Investors Consider MRVL Stock After the Decline?

The after-hours decline makes Marvell cheaper than it was before earnings, but cheaper does not automatically mean inexpensive.

Using the $241.45 regular-session close and Marvell’s Q3 basic share-count guidance of approximately 900 million produces a rough equity value of $217 billion.

Applying the 7.8% after-hours decline gives an indicative after-hours level of approximately $222.62 and a rough equity value of $200 billion. Extended-hours prices can change before regular trading opens.

This is an expectations test, not a complete valuation model. It does not fully account for cash, debt, preferred shares, warrant exercises, acquisitions or future share-count changes.

Still, the calculation shows that Marvell’s valuation continues to depend heavily on forecasts that stretch into FY2028 and FY2029. Investors are paying for the company to deliver the custom-chip ramps, expand operating margin and convert its AI opportunity into cash.

Our view is that Marvell’s operating story strengthened after these results, but the stock’s expectation risk did not disappear. The business demonstrated that connectivity remains a powerful current engine. The custom-silicon opportunity is credible, but its largest incremental contribution remains further away and less precisely quantified.

Potential investors can therefore frame the decision around evidence rather than the after-hours percentage:

• Is a 15% sequential revenue ramp in Q3 achievable?
• Can operating margin rise while gross margin declines?
• Will FY2029 guidance justify the expectations attached to Google?
• Are higher earnings reaching each share, or being diluted?
• Is the investor comfortable relying on FY2028 and FY2029 forecasts?

Different answers can reasonably lead to different conclusions. What would be risky is treating either the Google agreement or the 7.8% decline as a complete investment thesis.

Marvell Stock Q2 Earnings Analysis: Final Takeaway

This was not a broken quarter. It was an expectation reset. Marvell’s data-center engine passed the test. Q3 guidance passed. Operating leverage passed. Cash generation remained healthy. Management also raised FY2027 revenue to $12 billion and FY2028 revenue to $18 billion.

What failed was the size and timing of the surprise investors expected from Google. The market had begun treating the Google agreement like a new near-term revenue engine. Management clarified that it is better understood as validation of existing FY2028 expectations and a potentially much larger option for FY2029 and beyond.

That makes October 6 the next major checkpoint. Marvell must show that the future custom-chip story can produce not only enormous headline numbers, but profitable revenue, strong cash flow and attractive per-share growth.

Until then, the cleanest reading is this: Marvell’s business is accelerating, but the stock had accelerated even faster.