[DiaryPoint = Reporter Choi Seung-jin] Chinese memory maker CXMT said on October 30 that it had completed validation of an engineering sample for a sixth-generation DRAM process platform based on 4F². The announcement signals progress in product development, but it does not mean a product that could soon lower memory prices for smartphones or PCs is on the market.
4F² describes the area occupied by a single DRAM cell. F refers to the minimum feature size used in a process, so a 4F² cell has an area of four F squared. Assuming the same feature size as the widely used 6F² structure, the cell area could theoretically be about one-third smaller, potentially allowing more data to fit in the same space.
A structure that saves space by standing cells upright
DRAM stores data in tightly packed cells that hold an electrical charge. The 4F² design reduces the space needed for each cell by placing the transistor vertically. Smaller cells could help increase storage capacity per chip area or make chips with the same capacity smaller.
However, a smaller cell area does not mean the price of a finished product will fall by the same proportion. Memory chips also include peripheral circuits and input/output structures, while process complexity, yields, production volumes, and customer qualification all affect costs and supply. The theoretical density of a 4F² structure should be considered separately from the price of memory products consumers buy.
Announcement marks a sample-validation milestone
CXMT disclosed that validation of its engineering sample is complete. That is a meaningful step in development, but the company’s public materials do not make it possible to confirm specific capacity and speed, yield, customer evaluation results, or production scale. CXMT mentioned a gradual ramp-up, but gave no launch timeline or supply volume for consumer products.
Samsung Electronics also presented research on 4F² DRAM using vertical-channel transistors at the 2026 IEEE International Solid-State Circuits Conference (ISSCC). CXMT is not alone in pursuing next-generation cell designs. SK hynix is also researching future DRAM structures, but CXMT’s announcement alone is not enough to establish production-scale gaps or performance advantages among the companies.
For the announcement to affect prices of LPDDR for smartphones or memory for PCs, validated products would need to be adopted by customers and supplied in meaningful volumes. This news points to a technical advance at an earlier stage. CXMT’s future disclosures on product specifications, production timing, and supply scale will be key indicators of whether the consumer memory market is set to change.