The most striking element is the breadth of the growth. Revenue of $3.5 billion was up 19% from a year earlier and 10% from the prior quarter. By segment, automotive generated $1.94 billion (up 12%), industrial & IoT reached $755 million (up 38%), communications infrastructure and other climbed to $452 million (up 41%), and mobile rose 6% to $351 million. The key takeaway is that the recovery is not being driven by a single segment but is unfolding simultaneously across all regions and end markets.

Profitability metrics improved in tandem. The adjusted gross margin was 58.0%, the adjusted operating margin came in at 35.1%, and adjusted EPS of $3.61 was 33% higher than a year earlier. With profit growth (33%) running well ahead of revenue growth (19%), volumes are clearly flowing through to earnings above the fixed-cost base — a signal that the semiconductor cycle has entered an up-phase.

Cash generation was also solid. Operating cash flow reached $860 million, and adjusted free cash flow (after capex) totaled $791 million, representing 22.6% of revenue. Of this, $360 million was returned to shareholders through dividends and buybacks.

The negatives here do not mean the absolute results were poor — rather, the market had already priced in even better outcomes. Revenue exceeded estimates by roughly 1% and adjusted EPS by approximately 2.4%. For a stock that has rallied sharply during the semiconductor up-cycle, a beat of this magnitude tends to be read as merely "in line with expectations."

From a business mix perspective, the relative deceleration in automotive is the key point to watch. Automotive generated $1.94 billion, accounting for more than half of total revenue, yet its growth rate of 12% trails industrial & IoT (38%) and communications infrastructure (41%) by a wide margin. This means the early recovery momentum is coming from outside of automotive, and if automotive demand does not catch up, the company's overall growth rate could moderate over time.

Finally, there is a variable beyond the company's control. At the time of the release, the broader market was increasingly skeptical about whether AI infrastructure spending would translate into real returns, dragging down semiconductor stocks across the board. In an environment where strong individual results cannot overcome sector sentiment, it is not unusual for shares to fall even when the numbers are good.

Management's message is focused not on "results riding a cyclical recovery" but on "a phase where the structural growth axis is shifting." They frame AI as a trend that is moving beyond data centers into physical devices such as vehicles, factories, and robots, and argue that these are precisely the areas where the company has long held strength. The decision to highlight data centers as a new growth engine for the first time is also a notable shift.

The tone of the guidance is also close to confident. The Q3 outlook midpoint assumes high-single-digit sequential growth and roughly 20% YoY growth, while the lower end of the adjusted EPS guidance range is already above this quarter's actual EPS. In other words, management views this quarter not as a peak but as the middle of an up-phase. Even so, the market's cold reaction is being read as a judgment that this outlook had already been priced in.

Both results and guidance came in above expectations, yet the stock declined in extended trading. Two keys explain this divergence. First, the beat was thin. Revenue exceeding by 1% and adjusted EPS by just 2.4% is unlikely to provide fresh catalysts for a stock that had already largely priced in a semiconductor up-cycle. Second, the sector backdrop at the time of the release was unfavorable. As doubts grew over whether AI infrastructure spending would translate into actual earnings, semiconductor stocks across the board were under pressure, and strong individual results were not enough to reverse that tide.

The lesson for novice investors here is that "good earnings" and "a good stock reaction" are different problems. A stock does not compete with the numbers that are announced — it competes with the numbers the market was expecting before the announcement. That said, extended-hours moves tend to occur on thin volume and frequently reverse direction in the next regular session, so this decline should not be taken as a final verdict; it is better read alongside the flow of subsequent trading days.