Electroforming‐Free, Self‐Rectifying Selector‐Only Memory With Diffusive Cu‐Ion Dynamics for Logic‐In‐Memory Computing
ABSTRACT A selector‐only memory (SOM), performing the dual functions of a selector and memory in a single device, has emerged as a promising candidate to overcome critical issues such as high aspect ratio and thermal disturbance in 3D X‐point arrays (3DXP) for compute‐express‐link (CXL). Nevertheless, previously reported SOMs based on ovonic threshold switching (OTS) still have limitations in adoption as a new memory solution for 3DXP, owing to the requirement of an electroforming process at high voltage, bidirectional high current flow, and the intrinsic drift phenomenon. Here, we report a novel electroforming‐free self‐rectifying SOM (SR‐SOM) that exhibits a diode‐like unidirectional current flow and high endurance of ∼10 8 cycles. Previous SOM strategies have relied primarily on chalcogenide‐based OTS selectors to induce threshold voltage modulation. By contrast, our SR‐SOM operates via diffusive Cu‐ion dynamics, where the polarity‐induced threshold‐voltage shift originates from atomic migration under an applied bias, and the ultrafast self‐rupturing of conductive Cu filaments enables drift‐free properties with sub‐20‐ns. Moreover, by implementing a logic‐in‐memory scheme, we demonstrated all 16 Boolean logic functions and a full adder operation, verifying the multifunctional capability of SR‐SOM for both memory and logic operations.
Authors
- Jinsub Park (ORCID: https://orcid.org/0000-0003-1079-5532)
- Gwang‐Ho Park
- Jae-Kyeong Kim (ORCID: https://orcid.org/0000-0001-9486-2434)
- Dae-Seong Woo (ORCID: https://orcid.org/0000-0001-7301-580X)
- In‐Taek Woo
- Jea‐Gun Park
- Hong‐Gwang Seo
- Min‐Jong Han
Institutions
- SK Group (United States) (US)
- Seoul Semiconductor (South Korea) (KR)
- Advanced Materials and Devices (United States) (US)
- Hanyang University (KR)
Publication Details
- Journal
- Advanced Science
- Published
- 2026-08-25
- DOI
- https://doi.org/10.1002/advs.77423
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Ministry of Trade, Industry and Energy
- Samsung