Memory executives predict RAM shortage will last until 2028

Memory executives have forecast that the global shortage of RAM and other memory chips will persist through 2028, a prediction that signals a prolonged period of elevated costs for consumer electronics and data centers. This outlook, reported by Ars Technica, indicates that the tight supply conditions that began in late 2024 are not a temporary blip but a structural shift in the industry. Executives emphasized that prices for memory sold in 2027 are already significantly higher than those set for 2026, suggesting that the market has entered a new pricing baseline rather than a transient spike. This determination challenges previous assumptions that demand for high-bandwidth memory would normalize once initial production ramp-ups completed.

The root of this sustained shortage lies in a complex interplay of supply chain constraints and an unexpected surge in demand from artificial intelligence infrastructure. For decades, the memory market operated on a cyclical boom-and-bust model, with frequent overproduction leading to price crashes. However, the current cycle is being distorted by the massive capital expenditures required to build AI data centers. These facilities require vast quantities of high-bandwidth memory to support large language models and other generative AI workloads. Simultaneously, major manufacturers such as Samsung, SK Hynix, and Micron have strategically shifted their production capacity toward high-margin AI-specific chips, reducing the output of standard DRAM used in laptops, smartphones, and servers. This reallocation has created a scarcity for conventional memory, driving up prices across the board as manufacturers prioritize the most lucrative contracts.

The implications of this trend are already visible in the retail and enterprise sectors. Consumer hardware makers, including PC and smartphone manufacturers, are facing increased bill of materials costs that they are likely to pass on to end-users. Analysts expect this to result in higher sticker prices for new devices or, alternatively, a reduction in the base memory configurations offered in mid-range products. For enterprise clients, the cost of refreshing server fleets is rising, which complicates long-term budgeting for cloud providers and corporate IT departments. The shortage is not merely a matter of unit availability; it is affecting the lead times for delivery, with some industrial clients reporting wait times extending into several quarters. This friction in the supply chain forces companies to hold larger inventories, tying up capital that might otherwise be deployed elsewhere.

Looking ahead, industry experts suggest that a true resolution to the shortage is unlikely before 2028, as new fabrication plants come online and existing lines are retooled to handle standard memory production. The timeline for relief depends heavily on the continued growth of AI demand and the pace at which manufacturers can expand capacity without triggering a traditional oversupply crash. Until then, the memory market will remain in a state of tight equilibrium, with prices staying elevated and supply remaining constrained. This environment represents a significant departure from the historical volatility of the chip industry, marking a period where memory scarcity becomes a persistent factor in technology planning. Companies across the spectrum, from startups to global tech giants, must now plan for higher memory costs as a permanent feature of their operational budgets rather than a temporary market anomaly.