Micron Targets 100K HBM Wafers to Rival Samsung & SK Hynix
Micron Targets 100K HBM Wafers to Rival Samsung & SK Hynix
Micron is significantly scaling up its high-bandwidth memory (HBM) production capacity to meet artificial intelligence demand and close the competitive gap with its South Korean rivals. By the end of 2026, the company plans to add up to 60,000 wafer starts per month, bringing its total HBM capacity to approximately 100,000 wafers per month. This aggressive expansion effectively doubles its previous monthly output of roughly 40,000 to 50,000 wafers. To achieve this rapid scaling, Micron is routing new manufacturing equipment to its advanced packaging facilities in Taiwan and Singapore, while simultaneously preparing its Hiroshima plant in Japan for further equipment investments.
This aggressive production target is a strategic maneuver to challenge the market dominance currently held by SK Hynix and Samsung Electronics. Despite being a top-tier manufacturer, Micron has historically held the third-place position, securing an 18% market share in the second quarter of 2026 compared to SK Hynix’s 50% and Samsung’s 32%. The industry leaders currently operate at a massive scale, with estimated capacities ranging from 150,000 to 200,000 wafers per month—roughly three to four times Micron's previous scale. Reaching the 100,000-wafer milestone will allow Micron to significantly narrow this structural capacity gap and capture a larger slice of the highly lucrative AI memory market.
Beyond sheer volume, Micron is actively restructuring its product mix to prioritize next-generation memory architectures. While the company's current output remains largely dominated by 12-layer HBM3E products, it is accelerating the volume ramp of its advanced 12-layer HBM4 modules. Management projects that 12-layer HBM4 products will account for up to 50% of the company's total HBM mix by the end of 2026, rising sharply from a 20% to 30% share at the beginning of the year. This rapid technological pivot is specifically designed to supply the complementary memory required for next-generation AI accelerators, such as Nvidia's upcoming architectures.
The urgency behind this capacity build-out is underscored by an intense, sustained shortage in the global high-bandwidth memory supply chain. As AI chipmakers continue to scale up their hardware procurement for complex generative AI training systems, high-performance memory has become a critical component in short supply. Market analysts suggest that these tight supply-demand conditions will likely persist through 2028, with Micron already noting its capacity is 75% booked through 2027. By aggressively expanding its production footprint and successfully migrating to HBM4, Micron is positioning itself to convert this secular AI demand into sustained, long-term revenue growth.
Sources
Micron doubling HBM capacity
Micron targets 100K HBM wafers
Memory Supply Tight
AI Memory Boom Lasts Longer
Comments