存储疯狂紧缺!韩媒:三星电子将一季度NAND价格上调100%,而此前DRAM上调了70%
智通财经网·2026-01-25 04:00

Core Viewpoint - The global semiconductor market is experiencing severe supply-demand imbalances, leading Samsung Electronics to implement aggressive pricing strategies for NAND flash memory, with prices increasing by over 100% in Q1 2023, surpassing market expectations [1][2]. Group 1: Pricing Strategies - Samsung Electronics has completed negotiations with major clients for NAND supply contracts, officially implementing a new pricing system starting January 2023 [1]. - The price increase follows a nearly 70% rise in DRAM memory prices, indicating a significant shift in the storage market [1]. - Other major players, including SK Hynix, are also adopting similar pricing strategies, reflecting strong bargaining power in a seller's market [3]. Group 2: Demand Drivers - The demand for high-performance storage devices is being driven by the rapid expansion of AI infrastructure, particularly in enterprise solid-state drives (eSSD) and mobile devices requiring higher capacity storage for on-device AI applications [2][4]. - The market is witnessing exponential growth in demand due to the necessity for advanced storage solutions in smartphones and PCs to support local AI processing [4]. Group 3: Supply Constraints - The supply side is constrained by rigid production capacity, with major manufacturers, including Samsung, maintaining a cautious investment approach over the past year, resulting in limited growth in shipment volumes [4]. - There has been no large-scale capacity expansion in the NAND flash sector, and the industry's consensus indicates that effective supply has not kept pace with the surge in demand [4]. Group 4: Market Impact - The volatility in storage chip prices is becoming a bottleneck for AI industry development, significantly increasing the construction costs of AI infrastructure [4]. - The rising costs of DRAM and NAND are expected to be passed on to consumers, leading to higher prices for smartphones and PCs as manufacturers adjust to increased memory costs [4].