SK Hynix $1 Trillion Valuation Linked to AI Demand

SK Hynix Inc.’s equity market capitalization briefly exceeded $1 trillion on May 26, 2026, joining Micron Technology Inc. and Samsung Electronics Co. Ltd. in an intraday re-rating driven by surging AI data-center demand and constrained advanced-memory supply that is lifting high-bandwidth memory (HBM) prices.

Trillion-Dollar Milestones

On May 26, Micron’s market value topped $1 trillion for the first time, while SK Hynix’s intraday market cap surpassed $1.1 trillion. Samsung had crossed the $1 trillion threshold shortly before, placing three major memory suppliers at comparable valuation levels. No earnings releases, formal guidance updates, M&A announcements, or regulatory filings from SK Hynix, Micron, or Samsung within the prior 72 hours referenced these milestones, indicating the moves were driven by market dynamics rather than company actions. SK Hynix’s market cap rose from about $100 billion roughly 16 months ago to the intraday peak, with shares gaining about 11.1% in Seoul during the session.

AI Memory Demand and Outlook

Industry analysis attributes the re-rating to rising AI data-center demand and tight supply of leading-edge memory rather than near-term corporate events. AI workloads are straining memory bandwidth, shifting focus to stacked memory architectures placed on or near processors. Demand for HBM, a type of stacked memory used in AI accelerators, is projected to increase by more than 40% in 2026 alone, intensifying competition among suppliers.

Broader semiconductor forecasts based on World Semiconductor Trade Statistics (WSTS) project overall market growth of about 22% in 2025 and 26% in 2026. The memory segment is expected to expand roughly 28% in 2025 and 39% in 2026, supporting expectations for structurally higher pricing and volumes at memory suppliers. Upstream AI demand is significant, with a major AI-accelerator vendor reporting 114% revenue growth in 2024 and guiding for 63% growth in 2025. Industry estimates suggest global AI infrastructure funding could reach about $8 trillion from 2025 to 2029.

At the same time, semiconductor research warns of a possible AI-linked correction within the next year or two, signaling cyclical downside risk to current valuations. Sustained HBM demand and constrained supply could support structurally higher prices and volumes for major memory suppliers, but valuations remain sensitive to near-term shifts in demand or capacity.