SK hynix Unveils 3D FeNAND for AI Computing

What happened

SK hynix disclosed its research posted in a newsroom on Oct. 14, developing analog computing-in-memory (A-CiM) using ferroelectric three-dimensional NAND (3D FeNAND) to perform AI computations directly within NAND flash memory. The technology distinguishes current states by adjusting the number of programming pulses at the cell level, introduces air gaps between wordlines to reduce wordline resistance at the non-cell circuit level, and applies in-page pipelining (IPP) at the page level to optimize data placement and computation sequences. Simulations showed that 3D FeNAND-based A-CiM achieved up to 20 times higher throughput and 5,000 times greater energy efficiency than conventional nonvolatile memory.

Impact and signal

Affected components and suppliers

Source and time

Source: The Elec · 10/15/2026 · View original source