Research × Technology × Innovation
Magi Chen
Researcher | Technology Practitioner | Founder of GradInnovate
I am Magi Chen. With more than two decades of industry experience and a Ph.D. in Computer Science from National Tsing Hua University, I explore the connections between research and practical applications. My interests span electronic design automation (EDA/AI4EDA), Agentic AI, reinforcement learning, embedded Linux, and drone systems, alongside hypergraph partitioning, generative models, GPU-accelerated computing, and deep learning. As the founder of GradInnovate, I work across research, technology, and innovation, and welcome collaboration and exchange.
I also currently serve as an Assistant Professor in the Department of Computer Science and Information Engineering at National Changhua University of Education.
Research Focus
Research and technical interests across electronic design automation, intelligent agents, and embedded and autonomous systems.
-
Electronic Design Automation (EDA/AI4EDA)
Combining AI and electronic design automation to explore circuit design and optimization
-
Hypergraph Partitioning
Creating novel generative models for efficient VLSI design
-
Deep Learning
Applying neural networks to complex engineering problems
-
Agentic AI
Exploring intelligent agent design, reasoning, planning, and tool coordination
-
Reinforcement Learning
Exploring learning through environmental interaction and sequential decision-making
-
Embedded Linux
Exploring embedded Linux platforms, system integration, and hardware–software collaboration
-
Drone Systems
Exploring the integration of drone platforms, autonomous systems, and intelligent applications
Selected Publications
All publications-
HyperPart: A Multilevel Hypergraph Partitioner with Dual-Path Variational Embeddings
45th IEEE/ACM International Conference on Computer-Aided Design (ICCAD) · 2026
-
HyperPlace: Harnessing a large language model for efficient hyperparameter optimization in GPU-accelerated VLSI placement
ACM Transactions on Design Automation of Electronic Systems · 2025
-
GPart: A GNN-enabled multilevel graph partitioner
In Proceedings of Design Automation Conference · 2025
-
GenPart 2.0: Enhanced hypergraph partitioning with vertex weight handling using a generative model
ACM Transactions on Design Automation of Electronic Systems · 2025
Teaching
View courseInterested in research collaboration or industry applications?
Get in touch about research collaboration, industry applications, or teaching.