Samsung Electronics will invest roughly $648 billion over 10 years, allocating 300 trillion won for new memory chip fabs with some projects accelerated from the 2040s to mid-2030s. This capacity expansion could alleviate memory bottlenecks affecting Nvidia’s GPU production and potentially lower DRAM and flash prices.
Samsung Electronics has unveiled a 10-year investment blueprint totaling approximately $648 billion, with 300 trillion won earmarked for constructing and expanding memory chip fabrication plants. Projects once scheduled for the 2040s are now slated for completion by the mid-2030s, reflecting urgent demand for memory driven by AI and data center growth.
The planned memory capacity boost aims to ease chronic DRAM and NAND shortages that have constrained GPU output and inflated component costs. Expanded production could enable Nvidia to scale GPU shipments more rapidly and improve gross margins if component pricing stabilizes or declines.
South Korea’s presidential office has steered the plan to address regional development goals and secure electoral support, pushing investments beyond traditional Seoul-area clusters. The political mandate to diversify chip plant locations has accelerated timelines and broadened the geographic footprint of semiconductor manufacturing.
