Biristor coupled oscillators with biristor and oscillatory neural network using the same
The biristor-based oscillating neural network addresses the limitations of conventional oscillators by using silicon semiconductor-based biristors, achieving compatibility with semiconductor processes and enabling miniaturization while solving the Vertex Coloring Problem and functioning as an Ising Machine.
Patent Information
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- KOREA ADVANCED INST OF SCI & TECH
- Filing Date
- 2024-07-02
- Publication Date
- 2026-07-21
AI Technical Summary
Conventional oscillatory neural networks using insulator-metal transition materials or magnetic tunnel junction-based oscillators face limitations due to large and complex circuits, and incompatibility with existing semiconductor processes, leading to high power consumption and integration challenges.
A biristor-based coupled oscillator and oscillating neural network utilizing silicon semiconductor-based biristors, which are compatible with existing semiconductor processes and enable miniaturization, allowing phase synchronization of oscillating signals through a coupling capacitor.
The solution provides high compatibility with semiconductor processes, enables miniaturization, and can solve the Vertex Coloring Problem by converging phase differences to a predetermined value, functioning as an Ising Machine.
Smart Images

Figure 112024071813763-PAT00003_ABST