DIFFERENTIAL IONIC ELECTRONIC TRANSISTORS

Differential ionic transistors with coupled charged void and ion intercalation mechanisms address energy and reliability issues in machine learning algorithms, providing efficient, temperature-independent computation for embedded systems.

DE102021208846B4Active Publication Date: 2025-07-31ROBERT BOSCH GMBH
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
DE102021208846
Authority / Receiving Office
DE · DE
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-08-26
Filing Date
2021-08-12
Publication Date
2025-07-31
Estimated Expiration
2041-08-12

AI Technical Summary

Technical Problem

Existing computing devices for implementing machine learning algorithms are energy-intensive and unsuitable for low-power embedded systems, facing limitations such as accuracy in programming memory levels, temperature stability, and data transfer constraints, which affect their performance and reliability in harsh environments.

Method used

Differential ionic transistors with a coupled charged void formation mechanism and ion intercalation/deintercalation mechanism for accurate conductance and resistance control, operating in a temperature-independent manner, reducing power consumption and enhancing computational efficiency.

Benefits of technology

The ionic transistors provide multi-state temperature-independent switches with high power efficiency, enabling efficient computation in embedded systems by counting ionic charge rather than applying voltage or current, thus overcoming energy and reliability constraints.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

An ionic transistor (10) comprising:a first layer (14) comprising a first source contact (S1) and a first drain contact (D1), the first drain contact (D1) being spaced from the first source contact (S1), and a channel (38) extending between the first source contact (S1) and the first drain contact (D1);a first storage layer (42) electrically connected to the first source contact (S1) and the first drain (D1), the first storage layer (42) being located in the channel (38) and being formed from an electrically conductive material configured to conduct ions;a second layer (12) comprising a second source contact (S2) and a second drain contact (D2), wherein the second drain contact (D2) is spaced from the second source contact (S2), wherein the second source contact (S2) is spaced from the first source contact (S1), and wherein the channel (38) extends between the second source contact (S2) and the second drain contact (D2); a second storage layer (40) electrically connected to the second source contact (S2) and the second drain contact (D2), wherein the second storage layer (40) is located in the channel (38) and is formed from an electrically conductive material configured to conduct ions;andan electrolyte layer (14) located between and electrically connected to the first and second storage layers (42, 40), the electrolyte layer (14) consisting of a solid electrolyte (44), the solid electrolyte (44) being an insulator configured to prevent the passage of electrons but allow the passage of ions;
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Electrode, method of producing the same, battery, and electronic device

    US20160344032A1

  • Battery-based neural network weights

    US20180260681A1

  • Three-terminal non-volatile multi-state memory for cognitive computing applications

    US20180261764A1

  • Multilayer memristive devices

    US8471234B2