Enhancing Structure Functions for Accurate Thermal Characterization and Monitoring of Semiconductor Packages: Sampling Optimization and Geometric Analysis
본문
- Conference
- 제 32회 한국반도체 학술대회
- Date
- 2025-02-13
- Presentation Type
- 포스터
Abstract
As semiconductors become more highly integrated and high-performing, packages that physically protect the semiconductor while ensuring optimal performance has become increasingly important. Structure functions are commonly used for diagnosing degradation and optimizing designs, yet they face accuracy limitations, especially due to their reliance on transient temperature data sampling and assumptions of one-dimensional thermal structures.
This research addresses these limitations by identifying optimal sampling strategies that improve data efficiency and by analyzing structure function variations in three-dimensional shapes and aspect ratios to better represent complex heat flows in modern power semiconductor packages. Also, we propose two new metrics: the Surface Method for three-dimensional analysis and the Dynamic Time Warping (DTW) Method, to quantify and compare the structure function accuracy.
These methods aim to improve thermal characterization accuracy, supporting more precise design and on-line diagnostics of semiconductor packages.
- 다음글회전체 진동 신호 디노이징을 위한 노이즈 추정 및 확산 모델 기반 복원 네트워크 개발 24.11.06