Samsung and SK hynix Target Sub-10nm DRAM by 2028

According to industry sources cited by Korean media, Samsung Electronics and SK hynix are expected to advance DRAM manufacturing to sub-10nm process nodes around 2028, although the two companies are pursuing different memory cell architectures.

Samsung and SK hynix are reportedly targeting mass production of 1d DRAM in 2027, followed by limited initial production of 0a DRAM in 2028. In the DRAM naming scheme, 1d represents the seventh-generation 10nm-class DRAM, while 0a refers to the first generation of sub-10nm DRAM. The latest DRAM generation currently in mass production is 1c.

The timeline is later than the roadmap previously expected in 2025, when 1d production was projected for 2026 and 0a for the second half of 2027. The delay reflects not only technology readiness but also manufacturing economics. With DRAM prices remaining strong, memory makers have less incentive to accelerate costly process transitions that require equipment reconfiguration and yield stabilization.

Samsung is expected to introduce a vertical channel transistor (VCT) structure starting with the 0a node, while SK hynix is developing a vertical gate (VG) structure and is still evaluating whether to implement it at 0a or defer it to the following 0b generation.

Samsung has reportedly made progress with its new architecture, producing functional chips using a 4F² cell design and VCT structure on a sub-10nm process. The company is targeting mass production in 2028. Compared with the conventional 6F² DRAM cell, the 4F² design can reduce the cell footprint from 3F × 2F to 2F × 2F, potentially increasing memory density by approximately 30% to 50% within the same chip area.

The transition to smaller DRAM nodes is also expected to increase demand for advanced lithography. Samsung and SK hynix are reportedly considering ASML's High-NA EUV technology for 0a DRAM, initially applying it to only one or two of the most critical layers. The limited deployment could help balance higher resolution requirements, equipment costs, and early-stage production yields, as a High-NA EUV system is estimated to cost around $400 million, roughly twice the price of existing EUV equipment.