NeuroShine is a team led by Associate Professor Yu Chi-Hua from the Department of Engineering Science at National Cheng Kung University. Assemiconductor packaging enters the era of heterogeneous integration, the
team has developed an “advanced packaging EDA tool” that leveragesgenerative AI to simulate and emulate the environmental conditions of theentire chip system. Without the need to build massive datasets in advance,it can generate tailored solutions through a more advanced and accelerated computational model.

In the era of extremely complex advanced packaging, smarter AI design tools are essential.
-Original Technology: By integrating multi-scale simulation with artificial intelligence, we establish a bidirectional mapping between design and physical properties to tackle complex challenges.
-Core Technology: A unified system combining “simulation + generative AI”
that can autonomously generate its own database, enabling deep learning capabilities even without large datasets.
-It simulates how various physical characteristics behave across heterogeneous materials-such us coefficient of thermal expansion, thermal conductivity, surface adhesion, and warpage.

Trough reinforcement learning, it identifies 2D solutions that can be expanded into 3D, significantly reducing computational data and accelerating performance.

InPack.AI an AI-driven EDA tool that integrates generative AI for advanced packaging design faster, smarter, and easier to use.
InPack.AI enables heterogeneous integration, physical verification, and fast simulation-seamless with mainstream design software.