Key Highlights

  • Micron Technology (Nasdaq: MU) on the Nasdaq Global Select Market has experienced gross Margin expansion from approximately 22% at the FY2023 trough to 74% in Q2 FY2026 — a swing of 52 percentage points in approximately two and a half years.
  • DRAM average selling prices rose in the mid-110% range year-on-year in Q2 FY2026; historically, periods of pricing this strong in memory have been followed within 12-18 months by Supply additions that compress margins.
  • The company's FY2025 10-K states that memory Demand driven by AI is structurally increasing, with cloud servers requiring significantly increasing quantities of DRAM and NAND as AI workloads grow — a demand floor with no clear ceiling.
  • New supply from Micron's Idaho fabs will not reach the market until mid-2027 at the earliest; New York supply begins from 2030, creating a multi-year window during which structural demand may continue to outpace supply additions.
  • The risk Factor disclosed by Micron — that new fab construction by all three major DRAM producers could produce oversupply if not matched by demand — is the defining uncertainty in assessing cycle duration.

 

Every supercycle in semiconductor memory history has ended. The question is never whether a cycle will turn but when, and the answer to that question in the current environment is genuinely uncertain in a way that previous cycles were not. Micron Technology, Inc. (NASDAQ: MU) is the instrument through which this uncertainty can be measured most precisely, because its financial disclosures contain both the evidence for the cycle's durability and the risks that could terminate it.

What the History Shows

The memory industry has experienced several distinct pricing cycles over the past three decades, each with its own character but a common structure. Demand improves — typically driven by a new end-market adoption wave. Pricing recovers. All three major producers interpret the improved pricing as a signal to invest in new capacity. Capacity arrives, typically with a two-to-four year lag. Supply exceeds demand. Pricing collapses. Producers cut Capital Expenditure and write down inventory. The cycle repeats.

The FY2023 trough provides the most recent reference point. Micron's gross margins fell to approximately 22%, the company reported significant net losses, and capital expenditure was reduced. The recovery from that trough has been faster and more complete than any preceding cycle: gross margins reached 68% for H1 FY2026 and 74% in Q2 FY2026. DRAM average selling prices have risen in the mid-110% range year-on-year. Operating Income of $16.14 billion in Q2 FY2026 is, in dollar terms, approximately nine times the operating income of Q2 FY2025.

The New Variables

Two variables make the current cycle structurally different from its predecessors, and both are described in Micron's filings. The first is the nature of demand. The 10-Q states that AI-driven growth in the data centre has accelerated demand for memory and storage at a rate greater than the industry's ability to increase supply. The FY2025 10-K states that cloud servers supporting AI workloads require significantly increasing quantities of DRAM, including HBM, and NAND as AI becomes increasingly memory-centric. This is demand driven by the Investment/">Capital Investment decisions of companies with multi-year infrastructure commitments — Microsoft's $80 billion AI infrastructure plan for FY2025 was publicly disclosed; Amazon, Google, and Meta have made comparably large commitments. These are not consumer purchase decisions that can be deferred when memory prices rise.

The second variable is the product mix shift toward HBM. HBM is not a Commodity product. It cannot be manufactured at high Yield without leading-edge process technology and specialised advanced packaging capability. The three companies capable of producing HBM — Samsung, SK Hynix, and Micron — are all simultaneously facing a situation where HBM demand is outrunning their HBM-specific capacity. Expanding HBM capacity requires different investments than expanding commodity DRAM capacity, and the timeline is longer. This structural constraint on HBM supply is a ceiling on the rate at which the AI memory market can be served, regardless of how much conventional DRAM capacity is added.

The Supply Timeline

Micron's capacity addition timeline is disclosed specifically. The first Idaho fab will produce its initial DRAM wafers in mid-calendar 2027 — approximately 18 months from the time of this writing. The second Idaho fab will be operational by end-2028. The New York fabs will produce supply from 2030 onward. This timeline means that meaningful incremental DRAM supply from Micron's new US facilities is at least 18 months away, and the bulk of the new capacity is three to five years away. In the interim, growth in Micron's supply comes primarily from bit density improvements per wafer — the process node transitions that reduce per-bit Manufacturing costs — rather than from new wafer starts.

The combination of inelastic AI demand, a constrained HBM supply structure, and a multi-year lag between capex decisions and new supply means that the current supply-demand imbalance could persist longer than historical cycle analysis would suggest. Whether it does depends on variables that include the pace of AI infrastructure investment by hyperscalers, the rate at which Samsung and SK Hynix add HBM-specific capacity, and whether Chinese state-backed producers can eventually deliver meaningful DRAM supply to global markets.

Disclaimer: This article is for informational and educational purposes only and does not constitute financial, investment, or any other form of professional advice. All data and figures are sourced exclusively from Micron Technology Inc.'s (NASDAQ: MU) Form 10-Q for the quarter ended February 26, 2026 and Form 10-K for the fiscal year ended August 28, 2025, as filed with the US Securities and Exchange Commission.