AI-Driven Thermal Prediction for Enhanced Reliability in 3D HBM Chiplets
In this talk, Dr. Aladwani presents a data-driven methodology that leverages neural networks trained on finite-element simulations to predict junction temperatures and hotspot locations in 3D HBM chiplets. This approach significantly reduces simulation time while maintaining accuracy, enabling faster, reliability-aware design of advanced 3D IC systems.
About Dr. Tahani Aladwani:
Dr. Tahani Aladwani is a Research Assistant at the University of Glasgow. She earned her PhD in Computer Science (2024) and specializes in federated learning, data representation, and meta-learning techniques. Her research focuses on addressing challenges such as label inconsistencies, data imbalance, and improving model performance.
This talk was part of COIN3D, a Horizon Europe Twinning project, co-funded by the European Union.
Related Chiplet
- Integrated voltage regulator (IVR) chiplet
- High-performance connectivity chiplets
- eFPGA Chiplet
- DPIQ Tx PICs
- IMDD Tx PICs
Related Videos
- Use Cases for Chiplets in AI Clusters
- Machine Learning Applications in EDA for Chiplet Reliability
- Coding approaches for increasing reliability and energy efficiency of 3D technologies
- How to accelerate innovation in IC technology with chiplets and 3D ICs
Latest Videos
- 2026 DRAM and Advanced Packaging Master Class
- Chiplets, RISC-V, and Europe's AI Silicon Ambition
- Driving the Future of Automotive Compute with UCIe
- A Physics-Informed Neural Network Surrogate for Runtime PDN and Dynamic Droop Prediction in 2.5-D Chiplet Integration
- A Chiplet Interface Model for System-Level PPA Exploration