Skip to content
Sign in
SK hynix presents research on 3D FeNAND for analog compute-in-memory at IEEE IEDM 2025AI illustration
DevelopingTechnologyIndustryMarketsReported 2026-10-14 08:00

SK hynix presents research on 3D FeNAND for analog compute-in-memory at IEEE IEDM 2025

SK hynix researchers presented a paper at the 2025 IEEE International Electron Devices Meeting on co-optimizing cell, non-cell, and page schemes in 3D FeNAND to improve energy efficiency for analog compute-in-memory (A-CiM) applications. The work addresses AI computing bottlenecks by enabling computation within memory using ferroelectric NAND technology.

In depthRead the in-depth story

Why it matters. This advancement could reduce data movement in AI systems, improving efficiency for memory-intensive AI workloads by integrating storage and computation in a single architecture.

01

Who and what it touches

CXMTcompetes withSolidigmpart ofNVIDIApartners withmemory chipsmanufacturesCompute Express Link …usesjoint equipment devel…enablesNAVER Cloudagreement withstartupsinvests inSK hynix

Into SK hynix

From SK hynix

Each line is a relation with evidence behind it, drawn in the direction it points. A line shows that two things are connected, not how far this change reaches.

02

Evidence

  • SSK hynix NewsroomCompany2026-10-14 08:00
    Dividing a conventional 16KB page into four 4KB subpages enables data to be organized more flexibly according to the input and output sizes of neural network layers.
    View source