NVIDIA Gets the Headlines, But Micron Could Offer More Upside - Yahoo Finance

Amid the ongoing AI boom, NVIDIA remains at the center of market attention, with its CEO Jensen Huang driving investor interest through strategic partnerships in AI infrastructure, including collabora... Read original →

The AI compute race is shifting from logic chips to memory bottlenecks. While NVIDIA dominates the training chip ecosystem, surging HBM3E/HBM4 demand has made memory bandwidth critical—Micron now controls the 'data pipeline' of AI data centers by fully booking its 2026 HBM output. Technically, its collaboration with Marvell on CXL and near-memory computing could redefine server architecture. Geopolitically, tightening U.S. export controls push SK Hynix and Samsung to expand outside Taiwan, China, raising capex; Micron benefits from onshore manufacturing and subsidies, enhancing cost resilience. Intel may counter via IFS entering HBM packaging, while ST focuses on edge-AI memory solutions. Within 18 months, HBM will shift from premium option to standard in AI servers—Micron’s valuation could re-rate further if it cracks 3nm-class TSV stacking yields.