HOT RADAR
computeHEAT 11.2°

AI CURATED HOT STORY · 10/15/2026

SK Hynix Unveils 3D FeNAND Analog Compute-in-Memory: Leveraging Ferroelectric Polarization for Efficient AI MAC Operations

1 reports archived1 independent sourcesupdated 10/15/2026, 19:05:48
Synthesis & Latest Updates
1 Sources Cross-Validated

SK Hynix has published research on Analog Compute-in-Memory (A-CiM) using 3D FeNAND technology, aiming to reduce data movement and improve energy efficiency by performing MAC operations directly within memory. The approach leverages ferroelectric polarization states to store data and generate multiple current levels per cell, suitable for AI weights, marking a significant step in evolving non-volatile memory towards near/in-memory computing architectures.

LATEST/SK Hynix has published research on Analog Compute-in-Memory (A-CiM) using 3D FeNAND technology, aiming to reduce data movement and improve energy efficiency by performing MAC operations directly within memory. The approach leverages ferroelectric polarization states to store data and generate multiple current levels per cell, suitable for AI weights, marking a significant step in evolving non-volatile memory towards near/in-memory computing architectures.

HEAT TRENDHourly heat curve

3 fully observed hours
Now
11.2
Peak
11.820:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

TIMELINECoverage timeline

Total 1 reports · Latest first
  1. IT之家 · IT业界T2·68 pts
    • SK Hynix introduces 3D FeNAND technology integrated with Analog Compute-in-Memory (A-CiM) to minimize data movement overhead.
    • The core mechanism relies on ferroelectric polarization states for data storage, distinct from traditional NAND's charge-trapping structure.
    • FeNAND cells can generate multiple current states, naturally aligning with the storage and computation needs of AI model weights.
SK Hynix Unveils 3D FeNAND Analog Compute-in-Memory: Leveraging Ferroelectric Polarization for Efficient AI MAC Operations | Today for AI