Electronics procurement teams are facing renewed supply chain volatility as memory and storage prices climb in 2026. This is driven by sustained demand from AI data centres.
Since October 2025, contract pricing for both dynamic random-access memory (DRAM) and NAND flash has risen sharply, with some areas seeing double-digit percentage increases in successive quarters. This trend is altering supplier priorities, extending lead times, and moving allocation away from traditional consumer and embedded markets towards high-performance enterprise applications.
Memory and storage pricing trends
Market trackers report that NAND flash wafer contract prices climbed by between 20% and more than 60% in November 2025, as suppliers adjusted to stronger enterprise demand and tighter wafer availability. DRAM contract pricing similarly showed sustained growth, with some configurations rising by upwards of 15% month-on-month into late 2025.
The acceleration in pricing follows an extended period of inventory depletion at major memory vendors, where inventory levels for NAND and DRAM have fallen to historically low stock. That intensifies the effect of demand spikes on procurement costs and reinforces upward price momentum into early 2026.
AI data centres as primary demand driver
AI workloads demand memory and storage at higher capacities and bandwidths than many traditional enterprise applications. In particular, high-performance DRAM and high-capacity enterprise SSDs are central to data-centre architectures supporting large language models and other generative AI systems.
Industry research indicates demand for AI-optimised SSD configurations grew sharply in 2024 and 2025, and it is expected to continue its rapid expansion, with projected annual growth rates above 60% for AI-related SSD usage. This disproportionate growth channels available NAND and DRAM wafer capacity towards AI volumes, pressuring allocation for legacy segments.
Supply-side shifts and supplier strategy
Major memory suppliers have responded to these changing demand dynamics by prioritising production for high-margin, high-performance products and phasing down older, lower-density DRAM processes. This has reduced available capacity for commodity memory modules and lower-capacity storage typically used in consumer, industrial, and mid-tier embedded systems.
Longer lead times are now a common feature in memory contracts, with OEMs reporting delivery windows extending into mid-2026 for certain DRAM and NAND products.
Procurement implications
The current pricing environment has several consequences:
- Higher BOM cost pressures: rising contract prices for DRAM and NAND inflate BOM costs for systems that depend on memory and SSD capacity
- Extended lead times and allocation risk: as suppliers prioritise hyperscale and enterprise customers, lead times for commodity memory can extend significantly
- Shift in supplier negotiation dynamics: with memory suppliers reallocating capacity and tightening quotes, negotiation leverage may shift towards suppliers with available inventory or flexible fulfilment options
Broader market and supply-chain effects
The memory pricing surge is not isolated to server hardware. Independent analysis suggests that consumer electronics pricing, including PCs and smartphones, could be affected as material cost increases filter through to OEMs. Some forecasts indicate potential market contractions for devices in 2026 as memory cost pressures mount and supply remains constrained.
For many organisations, this has highlighted vulnerabilities in long supply chains where a small set of critical components – like DRAM and NAND – can dominate cost and availability risk when end-market demand shifts dramatically.
This current environment highlights the importance of data-driven forecasting, supplier engagement, and scenario planning that account for structural demand shifts, longer lead times, and allocation prioritisation. The memory supply chain appears poised for sustained volatility into 2026, largely driven by capacity realignment and evolving demand from AI infrastructure build-outs.
