EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Unveiling Search‐Robust One‐Shot Learning Operation in FeTFT‐Based Ternary Content‐Addressable Memory

Hyeonjung Park iD, Changwoo Han, Yejoo Choi, Junseok Kim, Myeongjae Choi, Huiseong Shin iD, Seungjoon Jeong, Jeongheon Cho, Chaeho Kwon, Sang Min Won iD, Changhwan Shin iD

Small Science · 2026-10 · 6 (10)

DOI: 10.1002/smsc.70421 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Small Science, online erschienen am 06.10.2026. Verfasst von Hyeonjung Park und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Electrical and Computer Engineering Sungkyunkwan University Suwon South Korea. Die Arbeit nennt 55 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Trade, Industry and Energy, National Research Foundation of Korea.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the rapid growth of data‐centric artificial intelligence (AI), ternary content‐addressable memory (TCAM) has emerged as a promising platform for parallel in‐memory search. Here, we systematically investigate how the gate‐dielectric material and structural design of metal–ferroelectric‐metal–insulator–semiconductor (MFMIS) ferroelectric thin‐film transistors (FeTFTs) affect TCAM search operation and long‐term robustness. By comparing ZrO 2 (11 nm), ZrO 2 (8 nm), and HfO 2 (8 nm) gate dielectrics, we decouple the effects of material substitution, thickness scaling, and equivalent‐oxide‐thickness (EOT)‐matched behavior. All devices show proper initial TCAM operation, with ZrO 2 8 nm exhibiting the highest initial sensing performance owing to its larger read current. Under repeated search cycling, however, ZrO 2 11 nm proves the most robust, while HfO 2 8 nm suffers the most severe degradation—governed by competition between electron trapping and donor‐like state creation rather than conventional ferroelectric fatigue. One‐shot classification simulations using experimentally extracted bit error rate (BER) trends confirm that device‐level degradation directly translates into accuracy loss, with ZrO 2 11 nm retaining the highest accuracy. Increasing the memcapacitor area ratio enlarges the memory window but degrades sensing response due to added parasitic capacitance and RC delay. These findings establish gate‐dielectric and interface design guidelines for search‐robust FeTFT‐based TCAM.

Schlagwörter

Stichwörter der Autor:innen

ferroelectric HZOferroelectric thin‐film transistorsin‐memory computingnonvolatile memoriesternary content‐addressable memories

Autor:innen und Institutionen

  • Department of Electrical and Computer Engineering Sungkyunkwan University Suwon South Korea.
  • School of Electrical Engineering College of Engineering Korea University Seoul South Korea.
  • Semiconductor R&D Center Samsung Electronics Hwaseong South Korea.
  • School of Semiconductor System Engineering College of Engineering Seoul South Korea.

ORCID-Profile: Hyeonjung Park, Huiseong Shin, Sang Min Won, Changhwan Shin

Förderung

  • Ministry of Trade, Industry and Energy
  • National Research Foundation of Korea

Zitieren

APA

Park, H., Han, C., Choi, Y., Kim, J., Choi, M., Shin, H., Jeong, S., Cho, J., Kwon, C., Won, S. M., & Shin, C. (2026). Unveiling Search‐Robust One‐Shot Learning Operation in FeTFT‐Based Ternary Content‐Addressable Memory. Small Science, 6(10). https://doi.org/10.1002/smsc.70421

Vancouver

Park H, Han C, Choi Y, Kim J, Choi M, Shin H, et al. Unveiling Search‐Robust One‐Shot Learning Operation in FeTFT‐Based Ternary Content‐Addressable Memory. Small Science. 2026;6(10). doi:10.1002/smsc.70421

BibTeX
@article{Park2026,
  author = {Park, Hyeonjung and Han, Changwoo and Choi, Yejoo and Kim, Junseok and Choi, Myeongjae and Shin, Huiseong and Jeong, Seungjoon and Cho, Jeongheon and Kwon, Chaeho and Won, Sang Min and Shin, Changhwan},
  title = {Unveiling Search‐Robust One‐Shot Learning Operation in FeTFT‐Based Ternary Content‐Addressable Memory},
  journal = {Small Science},
  year = {2026},
  volume = {6},
  number = {10},
  doi = {10.1002/smsc.70421}
}
Kostenlos starten →