Micron (MU) Earnings Analysis: Record Revenue of $4.14 Billion; SCA Strategy Reshapes Valuation Logic for the Memory Industry
TechFlow Selected TechFlow Selected
Micron (MU) Earnings Analysis: Record Revenue of $4.14 Billion; SCA Strategy Reshapes Valuation Logic for the Memory Industry
The storage supercycle is underway.
Navigating Web3 tides with focused insightsBy: Xiao Bing
One quarter: $41.4 billion in revenue—a staggering 346% year-on-year increase. Gross margin: 84.9%. Net income: $28 billion.
These figures would shock any tech company—but for Micron, they border on surreal. Just two years ago, in the same quarter, the company posted $9.3 billion in revenue, a 39% gross margin, and net income under $1.9 billion.
Yet its stock surged 15% after hours—not because of Q3 itself, but due to its Q4 guidance: $50 billion in revenue (midpoint), ~86% gross margin, and $31 earnings per share (EPS).
The memory supercycle is underway.
Beneath the Numbers: Dissecting a Money-Minting Machine
Breaking down Micron’s earnings report reveals exponential growth across every business segment.
DRAM generated $31.3 billion in revenue—76% of total revenue—with average selling prices rising over 60% sequentially. NAND delivered $9.9 billion, also far exceeding expectations. But the most explosive growth lies within specific business units: core data center revenue exceeded $25 billion this quarter—annualizing above $100 billion—up more than sevenfold from $1.53 billion a year earlier. Data center SSD revenue surpassed $5 billion, doubling sequentially. Automotive and embedded businesses recorded $4.63 billion, up over threefold year-on-year.
Operating cash flow stood at $25.39 billion; adjusted free cash flow reached $18.3 billion. Cash and investments totaled $30.2 billion; net cash amounted to $24.4 billion. Debt was reduced by $4.4 billion this quarter, and all three major rating agencies upgraded Micron to BBB+.
This earnings report renders the old adage—"memory is a cyclical industry"—obsolete.
16 SCAs: Micron Is Rewriting Its Business DNA
Numbers can be explained by supply-demand imbalances—but a sudden shift in business model cannot.
During its earnings call, Micron announced it has signed 16 Strategic Customer Agreements (SCAs). These are “take-or-pay” binding contracts spanning from 2026 through end-2030, covering approximately 20% of DRAM shipments and one-third of NAND shipments. The agreements include four hyperscale customers and three mid-sized customers, with the remainder coming from the automotive sector.
CFO Mark Murphy disclosed a new metric for the first time: Remaining Performance Obligations (RPO). As of the end of Q3, RPO stood at $5 billion; adding newly signed agreements post-quarter, this figure jumped to roughly $10 billion. Management explicitly stated that actual revenue will “far exceed” the contractual floor implied by RPO.
More critically, Micron aims to raise SCA-covered revenue to over 50% of total revenue.
What does this mean?
For the past 40 years, the memory industry’s business model relied on spot pricing and short-term contracts—price volatility was endemic, and valuation was persistently burdened by a “cyclical discount.” SCAs fundamentally transform memory chips from commodities into pre-sold infrastructure resources—identical in logic to cloud providers signing long-term contracts for power or fiber optics.
If Micron achieves >50% SCA coverage by 2027, its revenue predictability will resemble that of an enterprise software company—yet its forward P/E ratio stands at just over 10x. Among trillion-dollar market-cap companies, none is cheaper.
Supply Can’t Keep Pace With Demand
A single sentence from CEO Sanjay Mehrotra during the earnings call bears repeating: “Micron currently sees no point in time when supply can catch up to ever-growing demand.” He expects DRAM and NAND shortages to persist beyond 2027.
This is no idle claim. Micron’s entire 2026 HBM production capacity is already fully booked—priced and allocated. HBM4’s 12-layer stacking ramp is progressing at twice the speed of prior-generation HBM3E and has already contributed over $1 billion in revenue. Management forecasts the HBM market to grow at a ~40% CAGR—from $35 billion in 2025 to $100 billion in 2028—two years earlier than previously anticipated.
Supply-side constraints are physical. Building an advanced memory wafer fab takes 3–4 years and over $10 billion in investment. Micron has raised its FY2026 capital expenditures to ~$27 billion (net of government subsidies). New fabs in Idaho and Japan are ramping up, adding $100–200 million in startup costs each quarter. Yet capacity expansion is gradual, while AI data centers’ appetite for memory grows exponentially.
Global hyperscalers’ combined AI data center capex for 2026 exceeds $72.5 billion—money that will ultimately flow through memory chips.
The Timing of the Anthropic Deal Deserves Scrutiny
Two days before the earnings release, Micron announced a strategic partnership with Anthropic—including joint storage architecture design, multi-year supply agreements, on-site deployment of Claude at Micron, and a strategic investment in Anthropic’s Series H financing. With this, all three global HBM suppliers—Samsung, SK hynix, and Micron—are now strategic investors in Anthropic’s Series H round.
No AI lab has simultaneously secured supply commitments from all three global HBM manufacturers. This deal transcends mere supply contracts—it signals that AI companies now treat memory supply chains as strategic assets. As model training and inference efficiency becomes increasingly dependent on memory subsystem performance, memory vendors’ pricing power will only continue rising.
Concerns Remain—but Their Weight Is Shifting
The memory industry’s historical lesson is clear: every supercycle ends in a collapse triggered by overcapacity. Will Samsung, SK hynix, and Micron’s synchronized capacity expansions inevitably replay the crashes of 2018 or 2022?
The difference lies in structural demand shifts. Past memory cycles were driven by consumer electronics—once smartphone and PC shipments peaked, demand plummeted. But AI data center memory demand is cumulative: each new model grows larger; each inference consumes more tokens; each agent requires longer context windows. Add AI PCs standardizing memory from 16GB to 32GB and flagship smartphones concurrently increasing DRAM requirements—and the floor for memory demand has been structurally elevated.
The SCA framework serves as Micron’s institutional hedge against cycle risk. Even if demand growth slows, take-or-pay contracts guarantee a revenue floor. This won’t eliminate cycles—but it will dramatically narrow their amplitude.
Micron’s earnings season has ended—but the questions it raises have just begun: When memory evolves from a commodity into a pre-booked strategic resource, does the industry’s valuation language need rewriting?
Annualized on Q4’s $31 EPS guidance, Micron’s forward P/E sits around 10x. NVIDIA—similarly AI-benefited and supply-constrained—trades at over 30x forward P/E. That gap reflects both the market’s pricing-in of “cycles inevitably returning,” and an outdated narrative being incrementally disproven by the SCA framework.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
深潮TechFlow
