← Feed Deep Dive Matrix Subscribe

Watch Micron Gives Bullish Forecast, Warns on Margins - Bloomberg.com

www.bloomberg.com 2026-10-01 Bloomberg.com
Entities
Tags
MicronEarnings GuidanceMargin PressureMemory ChipsDRAMNAND FlashSemiconductorRevenue ForecastGross MarginStorageAI DemandData CenterHBMIndustry Competition
News Summary
Micron's latest guidance paints a nuanced picture of the memory chip industry at a critical inflection point. The bullish revenue outlook underscores how AI infrastructure buildout—particularly demand... Read original →
Industry Analysis
Micron's guidance is a head-on collision between AI demand narratives and the memory cycle's iron law. Technically, HBM's TSV stacking and hybrid bonding push per-GB costs 3 to 5x above commodity DRAM, yet Micron's yield ramp still trails SK Hynix by a full generation. Revenue growth is being cannibalized by depreciation and yield losses. On the NAND side, Samsung and Kioxia capacity additions hitting the market in H2 2025, combined with sluggish consumer recovery, put ASPs under 15 to 20 percent downward pressure—the direct source of the margin warning. Competitively, Samsung will almost certainly cut NAND pricing in Q3 to seize data center share while accelerating the HBM4 transition to rebuild its generational lead. SK Hynix, locked into NVIDIA's platform, will push pricing power further toward buyers, compressing industry-wide margins. Geopolitically, Micron's Idaho fab is constrained by immature local supply chains, while advanced packaging allocation sits with TSMC in Taiwan, China—a hidden bottleneck. Tighter export controls would accelerate mainland data center customers toward CXMT, structurally eroding Micron's NAND base. 12 to 24 month call: before HBM4 mass production in H1 2026, the industry enters a revenue-up-profit-flat squeeze. The margin inflection lives on the supply side, not demand—post-NAND clearance and HBM yield crossing 85 percent. Micron's guidance is a declaration: AI structural demand is real, but the cycle has not been repealed.
Read Original Article →
Related
This page displays AI-generated summaries and metadata for research purposes. Original content belongs to the respective publishers.