Core Insights - The article discusses the significant rise in memory prices, driven by both genuine demand from AI-related products and speculative behavior in the market [1][5][31] - It highlights the transition from a bear market in the storage sector to a speculative bubble, with memory prices reaching unprecedented levels, likened to real estate prices in major cities [1][13][32] Group 1: Memory Price Surge - The memory price surge began in 2025, with the price of 256G DDR5 server memory exceeding 40,000 yuan, and some high-end models reaching 49,999 yuan each, making a box of 100 units worth nearly 5 million yuan [1][12] - The price of consumer-grade DDR4 memory contracts surged by over 60% to 85% in Q3 2025, driven by panic buying and speculative hoarding by various stakeholders [12][28] - The memory market is experiencing a speculative bubble, with many believing that prices will continue to rise, despite the underlying demand being questionable [21][22] Group 2: Industry Dynamics - Major memory manufacturers like Samsung, SK Hynix, and Micron are shifting production capacity from consumer-grade products to high-margin AI storage products, which has led to a supply shortage in the consumer market [8][14] - The article notes that while AI has created real demand for memory, the underlying financial structures and speculative orders may not sustain long-term price increases [17][18] - The financial reports from companies like SK Hynix show that a significant portion of profits still comes from traditional DDR memory, indicating that the consumer electronics market remains crucial for profitability [23][25] Group 3: Speculative Behavior and Risks - The article describes a "nested doll" scenario where speculative orders for memory are based on unproduced chips and unbuilt data centers, creating a false sense of demand [18][19][20] - The potential for a market correction is highlighted, as the current high prices are based on optimistic assumptions about AI demand that may not materialize [30][31] - The article warns that if the anticipated AI-driven demand does not materialize, it could lead to significant losses for memory manufacturers and a broader crisis in the semiconductor industry [31][32]