Key Earnings Metrics

Micron's fiscal fourth quarter reflects a memory cycle dominated by AI infrastructure spending. DRAM, particularly high-bandwidth memory used alongside Nvidia and AMD accelerators, drove both revenue and profit higher.

The scale of the jump matters because it confirms that HBM pricing and volume are flowing through to the income statement, not just top-line growth. Below are the headline numbers from the quarter.

• Revenue nearly quadrupled YoY from $11.32 billion in the prior-year quarter, reflecting both higher HBM volumes and elevated memory pricing tied to constrained global supply.
• DRAM revenue reached $39.8 billion, up 343% year over year and representing 73% of total sales. This confirms DRAM remains the core profit engine, with HBM as the primary accelerator.
• Net income climbed to $37.7 billion, or $32.87 per share, compared with $3.2 billion, or $2.83 per share, a year ago. The margin expansion shows pricing power is translating directly into earnings.
• Data center revenue jumped 11-fold, underscoring that hyperscaler and AI accelerator demand, not consumer devices, is the dominant demand vector this cycle.
• Fiscal Q1 guidance came in at approximately $61.5 billion in revenue with adjusted EPS of $38.15, versus LSEG consensus of $35.40 EPS on $57 billion revenue. The guide above consensus implies Micron expects HBM tightness to persist into early fiscal 2026.
• Market cap topped $1.2 trillion, with the stock up more than 500% over the past year, reflecting investor confidence in sustained AI memory demand.

The combination of a large revenue beat and guidance above consensus tells us demand is not just strong but accelerating into the next quarter. The 11-fold data center growth is the most important structural number because it isolates AI infrastructure from consumer memory, which remains more cyclical.

Management Commentary

CEO Sanjay Mehrotra emphasized that Micron has a strong roadmap for future HBM products, positioning the company to compete more directly with SK Hynix and Samsung as capacity comes online. The roadmap framing matters because Micron currently holds the smallest HBM market share among the top three producers.

Mehrotra also highlighted collaboration with Nvidia on the industry's first custom HBM implementation. A custom HBM design win with Nvidia is strategically significant because it ties Micron more directly to the dominant AI accelerator supplier, potentially locking in multi-year demand beyond standard HBM allocations.

On workforce, Mehrotra said Micron increased compensation for every employee in fiscal 2026. This comes against a backdrop of labor tension, with hundreds of staffers at Taiwan factories threatening to strike over pay and labor disputes. Similar disputes at SK Hynix and Samsung have reportedly resulted in bonuses upward of $500,000, setting a benchmark Micron likely needs to meet to retain production-critical staff.

The compensation move is both a retention tool and a risk mitigation step. HBM production is capacity-constrained, and any work stoppage at a key facility could delay output at a moment when every wafer matters for meeting guidance.

Strategic Implications

Micron is investing $250 billion to build two new HBM campuses, a capital deployment scale that reflects how much supply-side pressure the company faces. The largest campus broke ground in Clay, New York, in January, and the first new fab in Boise, Idaho, is scheduled to come online next year.

The Boise fab coming online next year is the nearer-term milestone. It determines how quickly Micron can convert its roadmap and Nvidia collaboration into shipped HBM volume. Until then, the company is competing for share with a smaller production footprint than its Korean rivals.

SK Hynix and Samsung are also pursuing large HBM factory buildouts in South Korea, meaning Micron is racing against competitors who already hold larger market share and are expanding simultaneously. Micron's market cap exceeding $1.2 trillion shows investors are pricing in success, but execution on the Boise and Clay campuses will determine whether that valuation is sustained.

The global HBM supply crunch is the defining constraint. Advanced graphics and central processors from Nvidia and AMD require increasing amounts of HBM to handle AI workloads, and current production capacity across all three major suppliers is insufficient. This shortage has pushed memory costs higher, feeding into increased prices for consumer electronics including Apple's iPads and MacBooks.

That spillover into consumer pricing matters because it shows HBM tightness is not isolated to data center procurement. It is reshaping memory economics broadly, which could attract regulatory and political attention if consumer device inflation persists.

Mehrotra's recent participation in a White House AI regulation summit hosted by President Donald Trump, following a dinner with Chinese President Xi Jinping, places Micron squarely in the geopolitical conversation around semiconductor policy. For a company building multi-hundred-billion-dollar campuses in the United States, trade and export policy directly affects both demand access and equipment sourcing.

The labor risk in Taiwan is the most immediate operational threat. A strike at Micron's Taiwan factories could disrupt output at a time when HBM supply is already insufficient to meet demand. The compensation increases for fiscal 2026 are a direct response, but whether they fully resolve the disputes remains unresolved.

On the cycle question, Hendi Susanto at Gabelli Funds noted after the release that he had not heard negative data points pointing to a memory cycle reversal in the foreseeable future. That view aligns with Micron's above-consensus guidance, but memory cycles historically reverse quickly, and HBM capacity additions from all three major suppliers could eventually shift pricing power back to buyers.

What to Watch Next

I would watch four things over the next two quarters:

• Boise fab ramp timeline — The first new Boise fab is scheduled to come online next year, and any delay would directly limit Micron's ability to convert its Nvidia custom HBM collaboration into shipped volume.
• Taiwan labor resolution — Whether the compensation increases for fiscal 2026 fully resolve the threatened strikes will determine if production stays uninterrupted during peak HBM demand.
• HBM share trajectory — Micron holds the smallest HBM market share among SK Hynix, Samsung, and itself, so quarterly disclosures on HBM volume and customer mix will show whether the Nvidia partnership is closing that gap.
• Memory pricing signals — If DRAM and HBM pricing begins to flatten even as data center revenue stays elevated, it would mark the first sign that supply additions are catching up with AI-driven demand.

Related coverage

• AECOM Q2 2026: Segment Margin Hits 16.5%, Guidance Raised Again
• Equinix Q1 2026: Revenue $2.4B, AFFO Tops $1B, Guidance Raised
• Dell Q2 FY2026: $47B Revenue, $60.9B AI Server Orders, EPS $7.04
• Salesforce Q1 FY2027: Revenue $11.13B, Agentforce ARR Crosses $1B