Micron (MU) Q3 2025: HBM Revenue Run Rate Tops $6B as AI Memory Demand Accelerates

Micron’s Q3 saw record data center and HBM revenue, propelled by AI server demand and disciplined execution. Leadership is leaning into aggressive DRAM and NAND innovation, while CapEx and supply discipline remain central to margin expansion. With HBM4 sampling and a $200B US investment plan, Micron is positioning for sustained AI-driven growth into 2026 and beyond.

Summary

  • AI Server Pull-Through: Data center and HBM revenue strength reflect Micron’s deepening alignment with AI infrastructure buildout.
  • Technology Roadmap Execution: Rapid progress on 1Gamma DRAM and HBM4 sampling signal leadership in high-value memory segments.
  • Capital Commitment: $200B US investment plan and supply discipline underpin Micron’s long-term AI memory strategy.

Performance Analysis

Micron delivered a record $9.3B in revenue for Q3, up 15% sequentially and 37% year-over-year, with every major end market contributing to growth. Data center revenue more than doubled from the prior year, driven by surging demand for high-bandwidth memory (HBM) and AI-optimized DRAM products. The compute and networking segment reached $5.1B, with HBM revenue alone growing nearly 50% sequentially, reflecting Micron’s successful volume ramp and share gains in AI server platforms.

DRAM, which constituted 76% of total revenue, saw a 51% YoY increase, while NAND contributed 23% and grew 4% YoY. Gross margin expanded to 39%, up 110 basis points sequentially, as favorable product mix and disciplined pricing offset consumer-oriented pricing pressure. Operating income rose to $2.5B, and free cash flow exceeded $1.9B, a six-year high. Inventory days fell sharply, underscoring robust demand and improved supply chain management.

  • HBM Momentum: HBM revenue exceeded a $6B annualized run rate, with Micron on track to reach DRAM share parity ahead of schedule.
  • Segment Diversification: Mobile and embedded segments rebounded strongly, with mobile up 45% sequentially as inventory digestion subsided.
  • Margin Expansion Drivers: Mix shift toward higher-value DRAM and data center products continues to support gross margin improvement.

Micron’s performance was defined by AI-driven demand, operational discipline, and rapid technology execution, setting a strong foundation for Q4 and fiscal 2026.

Executive Commentary

"In fiscal Q3, DRAM revenue reached a new record driven by a nearly 50% sequential growth in HBM revenue. We remain the sole supplier in volume production of LPD RAM in the data center... Looking ahead to fiscal Q4, we see a robust demand environment and expect to grow revenue by 15% sequentially to a record $10.7 billion at guidance midpoint."

Sanjay Mehrotra, Chairman, President, and CEO

"Gross margins were above the high end of our guidance range, primarily due to better prices for both DRAM and NAND, partially offset by a higher consumer-oriented mix... We expect our revenue growth to be weighted towards DRAM, supported by robust pricing execution, favorable product mix, and continued cost improvements, all of which benefit gross margins."

Mark Murphy, CFO

Strategic Positioning

1. HBM and AI Memory Leadership

Micron’s HBM (High Bandwidth Memory) ramp is outpacing expectations, with the company now shipping high-volume HBM to four customers across GPU and ASIC platforms. The HBM3E and upcoming HBM4 nodes leverage internally developed logic dies and advanced DRAM, enabling >2TB/sec per stack and industry-leading power efficiency. The company expects HBM4 volume production in 2026, aligning with next-gen AI platform launches.

2. Technology Node Acceleration

Micron’s 1Gamma DRAM node, using EUV lithography, offers 30% better bit density, 20% lower power, and 15% higher performance than 1Beta. Early shipments of 1Gamma-based LP5 DRAM and a rapid yield ramp signal strong execution and future cost competitiveness. In NAND, the G9 node and QLC mix reached record highs, with new SSD products targeting both enterprise and consumer markets.

3. Capital and Supply Discipline

With a $200B US investment plan (including $150B for manufacturing and $50B for R&D), Micron is building two new Idaho fabs and expanding advanced packaging. CapEx remains focused on HBM and leading-edge DRAM, with fiscal 2025 spending unchanged at $14B. Notably, Micron is structurally reducing NAND wafer capacity by 10% to ensure capital efficiency and avoid oversupply, a key lever for margin stability.

4. Customer and Market Alignment

Micron’s reorganization around key market segments aims to deepen engagement with AI-focused customers. The company is now the number two brand in data center SSDs and holds sole-source status for LP server DRAM. Automotive and industrial segments are rebounding, with new product wins and channel inventory normalization supporting growth.

Key Considerations

Q3’s results underscore Micron’s successful pivot to AI-centric memory and storage, but also highlight the complexity of balancing aggressive investment with market and supply discipline.

Key Considerations:

  • HBM Supply and Share Gains: Micron expects to reach DRAM market share parity in HBM ahead of previous guidance, supported by rapid yield and capacity ramp.
  • Technology Transition Pace: The 1Gamma and G9 node ramps will shape Micron’s cost structure and competitive positioning through 2026.
  • NAND Capacity Management: Structural wafer reductions and measured node conversions are designed to prevent margin-dilutive oversupply.
  • Capital Allocation: $200B US investment plan will require careful execution to avoid capital inefficiency, especially as AI demand cycles evolve.
  • Tariff and Regulatory Volatility: Customer ordering patterns reflect some tariff pull-ins, but management views the impact as modest for now.

Risks

Micron faces risks from volatile AI demand cycles, potential overbuild in HBM capacity, and uncertainties around global tariffs and regulatory policy. Aggressive CapEx commitments could pressure returns if demand falls short or if competitors catch up technologically. Legacy DRAM and NAND pricing remains sensitive to mix shifts and broader macroeconomic headwinds.

Forward Outlook

For Q4, Micron guided to:

  • Revenue of $10.7B, plus or minus $300M
  • Gross margin of 42%, plus or minus 100 basis points
  • Operating expenses of approximately $1.2B
  • EPS of $2.50, plus or minus $0.15

For full-year 2025, management maintained CapEx guidance at $14B, with most spending directed to HBM and advanced DRAM nodes. Factors highlighted include:

  • Continued tight DRAM and reduced NAND inventories exiting fiscal 2025
  • Constructive demand environment, especially in AI and data center

Takeaways

Micron’s Q3 confirms a decisive shift toward AI-driven memory demand, with HBM and advanced DRAM nodes at the center of its growth strategy.

  • AI Platform Alignment: Data center and HBM outperformance is translating into record revenue and margin expansion, with strong customer pull-through.
  • Execution on Technology and Supply: Rapid node transitions and disciplined capacity management are supporting both top-line growth and margin resilience.
  • Watch for AI Demand Cycles: Investors should monitor the sustainability of AI server buildouts, HBM4 adoption pace, and Micron’s ability to balance CapEx with ROI.

Conclusion

Micron’s Q3 results reflect clear operational leverage from AI memory demand and disciplined execution on both technology and capital fronts. The company’s aggressive investment in US manufacturing and leadership in HBM position it to capitalize on the next wave of AI infrastructure, though risks around demand cyclicality and capital intensity remain.

Industry Read-Through

Micron’s results highlight the centrality of high-bandwidth memory and advanced DRAM in next-generation AI platforms, with implications for the entire semiconductor supply chain. Competitors lagging in HBM yield or node transitions risk losing share, while NAND suppliers must navigate price and capacity discipline to avoid margin erosion. Capital intensity and supply discipline are emerging as key differentiators, and the pace of AI server deployments will set the tone for memory pricing and investment across the industry.