Micron Technology’s fiscal Q2 results weren’t just good—they were historic, with revenue soaring 196% and gross margins crushing past 74%. CEO Sanjay Mehrotra explicitly links this performance to AI infrastructure, and a Q3 projection for 81% margins confirms this is not a peak but a new plateau, forcing a reevaluation of the entire memory sector’s cyclicality.
For years, Micron Technology was a pure-play on the brutal, cyclical memory market. Boom and bust cycles dictated its fortunes, and investors treated its stock as a high-risk, timing-dependent wager. That paradigm has just been shattered. The company’s fiscal 2026 second-quarter results, reported on March 18, 2026, didn’t merely beat estimates—they established all-time records for revenue, gross margin, earnings per share, and cash flow, propelled by insatiable demand for high-bandwidth memory (HBM) and DRAM used in artificial intelligence (AI) servers and infrastructure.
The numbers are staggering. Revenue reached $23.9 billion, a 196% year-over-year increase and 75% jump from the prior quarter. Adjusted earnings per share (EPS) soared to $12.20, up 155%. Even more telling was the gross margin, which skyrocketed to 74.4%—an expansion of 3,760 basis points from the year-ago quarter. This level of profitability was unthinkable for Micron just two years ago, a fact confirmed by the official earnings release and call transcript.
The AI Supercycle Is Not a Mirage; It’s Measurable in Margins
CEO Sanjay Mehrotra directly attributed this performance to AI, stating, “Micron set new records across revenue, gross margin, EPS, and free cash flow in fiscal Q2, driven by a strong demand environment, tight industry supply, and our strong execution.” The key phrase is “tight industry supply.” Unlike past cyclical spikes where supply eventually overwhelmed demand, the current HBM and advanced DRAM supply chain is structurally constrained due to the immense complexity of manufacturing. This creates a sustained sellers’ market.
The segment data reveals the source of this growth:
- Cloud Memory: $7.7 billion, up 163% year-over-year. This is the core AI business.
- Data Center (core): $5.7 billion, up 211%. This includes legacy DRAM but is now also AI-infused.
- Mobile & Client: $7.7 billion, up 245%. AI PCs and premium smartphones are a new growth vector.
- Automotive & Embedded: $2.7 billion, up 162%. Advanced driver-assistance systems (ADAS) demand more sophisticated memory.
The guidance for fiscal Q3 is what truly stunned the Street. Micron projects revenue of $33.5 billion (260% YoY growth) and an adjusted EPS of $19.15, with gross margins expected to hit 81% at the midpoint. Wall Street’s consensus had been for $23.3 billion in revenue and $10.77 in EPS, meaning Micron is signaling a performance that is an order of magnitude ahead of expectations, as documented in the detailed transcript.
Why the Market’s Skepticism Persists: A Valuation Conundrum
Paradoxically, despite these monumental results, Micron’s stock has not re-rated to the stratospheric valuations seen in some AI pure-plays. It trades at a forward P/E multiple below 13x. This disconnect stems from a deep-seated investor memory of the 2022-2023 memory downturn, when a glut of chips led to massive losses. The core investor thesis is: “This is just another cycle peak; the downturn is coming.”
However, the current cycle is fundamentally different. The demand is not for commodity memory but for specialized, high-value HBM and LPDDR5X memory that is physically more complex to produce and directly tied to the build-out of AI infrastructure. The “tight industry supply” cited by Mehrotra is not a temporary capacity constraint but a new industry structure. Investors are discounting the possibility that the AI-driven need for more memory per server and per device represents a permanent, structural increase in the industry’s average selling price and gross margin baseline.
Connecting the Dots: From Cyclical Commodity to Strategic AI Enabler
To understand the magnitude of this shift, one must compare it to the Nvidia story. Nvidia’s ascent was fueled by the AI compute revolution. Micron is experiencing its own parallel ascent, fueled by the AI memory revolution. Every powerful AI GPU requires stacks of HBM to operate efficiently. Without this high-speed memory, the computational power is bottlenecked. Micron, alongside Samsung and SK Hynix, now controls this critical chokepoint.
The historical context is crucial. For decades, memory was a low-margin, capital-intensive commodity business. Profitability was fleeting. The Q2 gross margin of 74.4% and projected Q3 margin of 81% are not anomalies; they are the new normal for a product type with immense barriers to entry and no near-term substitution risk. This transforms Micron’s financial model from one of volatile earnings to one of sustained, high-margin profitability, a point reinforced by management’s 30% dividend increase to $0.15 per share, a gesture of confidence in the durability of cash flow.
Risks to Monitor: Cyclical Ghosts and Competitive Dynamics
No analysis is complete without acknowledging risks. The primary risk is the eventual increase in HBM production capacity. If all three major manufacturers significantly expand output in 2027-2028, prices could soften. Additionally, while AI demand is robust today, a slowdown in enterprise AI spending or a failure to monetize AI applications at scale could curb demand sooner than expected.
Competition is fierce but currently collaborative. The market is an oligopoly, and all players are benefiting from the price surge. The risk is not price competition but operational missteps. Micron’s execution must remain flawless as it navigates the complex transition to next-generation memory nodes.
The Investment Thesis: A Second Act Built on Silicon
For investors, the takeaway is clear. Micron is delivering on a transformation that has been theorized for years but never this convincingly proven. The Q3 guidance of $33.5 billion in revenue implies full-year fiscal 2026 revenue could approach $100 billion, a milestone that places it among the largest semiconductor companies by revenue, not just memory. This is not a one-quarter pop driven by a single customer; it is a broad-based, multi-segment explosion.
The stock’s underperformance relative to its own fundamentals suggests the market is anchoring to the past. The investment opportunity lies in the re-rating that occurs when the consensus accepts that the memory cycle has been altered by AI permanently. The historical precedent for companies that break their cyclical mold is a significant, long-term upward revaluation. Micron’s 348% surge over the past year is a start, but the path to a sustainably higher multiple is just beginning if the AI-driven margin profile holds.
The data is unequivocal: record revenue, record margins, record guidance, and a management team explicitly framing this as a new era. For investors seeking exposure to the AI buildout beyond just the GPU, Micron represents the indispensable memory layer of that stack. The spectacular news from CEO Sanjay Mehrotra is that the “memory cycle” as we knew it is over. Welcome to the age of strategic memory.
For the fastest, most authoritative breakdown of how this shift impacts your portfolio and what to monitor next, onlytrustedinfo.com is your definitive source for actionable financial intelligence. Our expert team cuts through the noise to deliver the clarity you need to make informed decisions in real-time.
