Completed my PhD at Cornell University
Earned my PhD after completing research on large-scale graph learning and circuit applications.
Chenhui Deng is a Senior Research Scientist at NVIDIA, where he focuses on advancing large language models for chip design. His research lies at the intersection of machine learning and electronic design automation (EDA), with the goal of developing AI-driven solutions for next-generation chip design. Prior to joining NVIDIA, he earned his PhD from Cornell University.
PhD
Cornell University
Bachelor
Huazhong University of Science and Technology
* denotes equal contribution.
ACE-RTL: When Agentic Context Evolution Meets RTL-Specialized LLMs.
NSF Workshop on Agents for Chip Design Automation (Agent4Chip), 2026.
@inproceedings{deng2026acertl,
title={ACE-RTL: When Agentic Context Evolution Meets RTL-Specialized LLMs},
author={Chenhui Deng and Zhongzhi Yu and Guan-Ting Liu and Nathaniel Pinckney and Brucek Khailany and Haoxing Ren},
booktitle={NSF Workshop on Agents for Chip Design Automation},
year={2026}
}Artificial Intelligence-Assisted IC Design: Large language models and agentic systems for modern chip design.
IEEE Solid-State Circuits Magazine, 2026.
@article{deng2026aiassisted,
title={Artificial Intelligence-Assisted IC Design: Large language models and agentic systems for modern chip design},
author={Chenhui Deng and Chia-Tung Ho and Haoxing Ren},
journal={IEEE Solid-State Circuits Magazine},
year={2026}
}ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation.
ACM/IEEE Workshop on Machine Learning for CAD (MLCAD), 2025.
@inproceedings{deng2025scalertl,
title={ScaleRTL: Scaling LLMs with Reasoning Data and Test-Time Compute for Accurate RTL Code Generation},
author={Chenhui Deng and Yun-Da Tsai and Guan-Ting Liu and Zhongzhi Yu and Haoxing Ren},
booktitle={ACM/IEEE Workshop on Machine Learning for CAD},
year={2025}
}ChipAlign: Instruction Alignment in Large Language Models for Chip Design via Geodesic Interpolation.
ACM/IEEE Design Automation Conference (DAC), 2025.
@inproceedings{deng2025chipalign,
title={ChipAlign: Instruction Alignment in Large Language Models for Chip Design via Geodesic Interpolation},
author={Chenhui Deng and Yunsheng Bai and Haoxing Ren},
booktitle={ACM/IEEE Design Automation Conference},
year={2025}
}Polynormer: Polynomial-Expressive Graph Transformer in Linear Time.
International Conference on Learning Representations (ICLR), 2024.
@inproceedings{deng2024polynormer,
title={Polynormer: Polynomial-Expressive Graph Transformer in Linear Time},
author={Chenhui Deng and Zichao Yue and Zhiru Zhang},
booktitle={International Conference on Learning Representations},
year={2024}
}Less is More: Hop-Wise Graph Attention for Scalable and Generalizable Learning on Circuits.
ACM/IEEE Design Automation Conference (DAC), 2024.
@inproceedings{deng2024hoga,
title={Less is More: Hop-Wise Graph Attention for Scalable and Generalizable Learning on Circuits},
author={Chenhui Deng and Zichao Yue and Cunxi Yu and Gokce Sarar and Ryan Carey and Rajeev Jain and Zhiru Zhang},
booktitle={ACM/IEEE Design Automation Conference},
year={2024}
}GARNET: Reduced-Rank Topology Learning for Robust and Scalable Graph Neural Networks.
Learning on Graphs Conference (LoG), 2022. Spotlight
@inproceedings{deng2022garnet,
title={GARNET: Reduced-Rank Topology Learning for Robust and Scalable Graph Neural Networks},
author={Chenhui Deng and Xiuyu Li and Zhuo Feng and Zhiru Zhang},
booktitle={Learning on Graphs Conference},
year={2022}
}SPADE: A Spectral Method for Black-Box Adversarial Robustness Evaluation.
International Conference on Machine Learning (ICML), 2021.
@inproceedings{cheng2021spade,
title={SPADE: A Spectral Method for Black-Box Adversarial Robustness Evaluation},
author={Wuxinlin Cheng and Chenhui Deng and Zhiqiang Zhao and Yaohui Cai and Zhiru Zhang and Zhuo Feng},
booktitle={International Conference on Machine Learning},
year={2021}
}GraphZoom: A Multi-Level Spectral Approach for Accurate and Scalable Graph Embedding.
International Conference on Learning Representations (ICLR), 2020. Oral
@inproceedings{deng2020graphzoom,
title={GraphZoom: A Multi-Level Spectral Approach for Accurate and Scalable Graph Embedding},
author={Chenhui Deng and Zhiqiang Zhao and Yongyu Wang and Zhiru Zhang and Zhuo Feng},
booktitle={International Conference on Learning Representations},
year={2020}
}Accurate Operation Delay Prediction for FPGA HLS Using Graph Neural Networks.
IEEE/ACM International Conference on Computer-Aided Design (ICCAD), 2020.
@inproceedings{ustun2020accurate,
title={Accurate Operation Delay Prediction for FPGA HLS Using Graph Neural Networks},
author={Ecenur Ustun and Chenhui Deng and Debjit Pal and Zhijing Li and Zhiru Zhang},
booktitle={IEEE/ACM International Conference on Computer-Aided Design},
year={2020}
}Agentic Hardware Design as Repository-Level Code Evolution.
arXiv preprint, 2026.
@misc{yu2026agentic,
title={Agentic Hardware Design as Repository-Level Code Evolution},
author={Cunxi Yu and Chenhui Deng and Nathaniel Pinckney and Brucek Khailany},
howpublished={arXiv preprint},
year={2026}
}GRPO with State Mutations: Improving LLM-Based Hardware Test Plan Generation.
arXiv preprint, 2026.
@misc{kochar2026grpo,
title={GRPO with State Mutations: Improving LLM-Based Hardware Test Plan Generation},
author={Dimple Vijay Kochar and Nathaniel Pinckney and Guan-Ting Liu and Chia-Tung Ho and Chenhui Deng and Haoxing Ren and Brucek Khailany},
howpublished={arXiv preprint},
year={2026}
}AUTOGATE: Automated Clock Gating via Toggling-Aware LLM-based RTL Rewriting.
arXiv preprint, 2026.
@misc{wang2026autogate,
title={AUTOGATE: Automated Clock Gating via Toggling-Aware LLM-based RTL Rewriting},
author={Yiting Wang and Chenhui Deng and Chia-Tung Ho and Yanqing Zhang and Zhuo Feng and Cunxi Yu and Ang Li and Gang Qu and Brucek Khailany},
howpublished={arXiv preprint},
year={2026}
}How LLMs Fail and Generalize in RTL Coding for Hardware Design?.
arXiv preprint, 2026.
@misc{liu2026howllms,
title={How LLMs Fail and Generalize in RTL Coding for Hardware Design?},
author={Guan-Ting Liu and Chao-Han Huck Yang and Chenhui Deng and Zhongzhi Yu and Brucek Khailany and Yu-Chiang Frank Wang},
howpublished={arXiv preprint},
year={2026}
}Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning.
arXiv preprint, 2026.
@misc{blakeman2026nemotron,
title={Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning},
author={Aaron Blakeman and Aaron Thomas and Aastha Jhunjhunwala and others},
howpublished={arXiv preprint},
year={2026}
}Comprehensive Verilog Design Problems: A Next-Generation Benchmark Dataset for Evaluating Large Language Models and Agents on RTL Design and Verification.
arXiv preprint, 2025.
@misc{pinckney2025cvdp,
title={Comprehensive Verilog Design Problems: A Next-Generation Benchmark Dataset for Evaluating Large Language Models and Agents on RTL Design and Verification},
author={Nathaniel Pinckney and Chenhui Deng and Chia-Tung Ho and Yun-Da Tsai and Mingjie Liu and Wenfei Zhou and Brucek Khailany and Haoxing Ren},
howpublished={arXiv preprint},
year={2025}
}SmoothE: Differentiable E-Graph Extraction.
ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2025. Best Paper
@inproceedings{cai2025smoothe,
title={SmoothE: Differentiable E-Graph Extraction},
author={Yaohui Cai and Kaixin Yang and Chenhui Deng and Cunxi Yu and Zhiru Zhang},
booktitle={ASPLOS},
year={2025}
}HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization.
International Conference on Learning Representations (ICLR), 2026.
@inproceedings{chen2026heurigym,
title={HeuriGym: An Agentic Benchmark for LLM-Crafted Heuristics in Combinatorial Optimization},
author={Hongzheng Chen and Yingheng Wang and Yaohui Cai and Hins Hu and Jiajie Li and Shirley Huang and Chenhui Deng and Rongjian Liang and Shufeng Kong and Haoxing Ren and Samitha Samaranayake and Carla P. Gomes and Zhiru Zhang},
booktitle={International Conference on Learning Representations},
year={2025}
}Jarvis: A Multi-Agent Code Assistant for High-Quality EDA Script Generation.
arXiv preprint, 2025.
@misc{pasandi2025jarvis,
title={Jarvis: A Multi-Agent Code Assistant for High-Quality EDA Script Generation},
author={Ghasem Pasandi and Kishor Kunal and Varun Tej and Kunjal Shah and Hanfei Sun and Sumit Jain and Chunhui Li and Chenhui Deng and Teodor-Dumitru Ene and Haoxing Ren and Sreedhar Pratty},
howpublished={arXiv preprint},
year={2025}
}ASPEN: LLM-Guided E-Graph Rewriting for RTL Datapath Optimization.
ACM/IEEE Workshop on Machine Learning for CAD (MLCAD), 2025.
@inproceedings{zhang2025aspen,
title={ASPEN: LLM-Guided E-Graph Rewriting for RTL Datapath Optimization},
author={Niansong Zhang and Chenhui Deng and Johannes Maximilian Kühn and Chia-Tung Ho and Cunxi Yu and Zhiru Zhang and Haoxing Ren},
booktitle={ACM/IEEE Workshop on Machine Learning for CAD},
year={2025}
}Marco: Configurable Graph-Based Task Solving and Multi-AI Agents Framework for Hardware Design.
Symposium on VLSI Technology and Circuits (VLSI), 2025.
@inproceedings{ho2025marco,
title={Marco: Configurable Graph-Based Task Solving and Multi-AI Agents Framework for Hardware Design},
author={Chia-Tung Ho and Jing Gong and Yunsheng Bai and Chenhui Deng and Haoxing Ren and Brucek Khailany},
booktitle={Symposium on VLSI Technology and Circuits},
year={2025}
}Graph Learning at Scale: Characterizing and Optimizing Pre-Propagation GNNs.
Conference on Machine Learning and Systems (MLSys), 2025.
@inproceedings{yue2025graph,
title={Graph Learning at Scale: Characterizing and Optimizing Pre-Propagation GNNs},
author={Zichao Yue and Chenhui Deng and Zhiru Zhang},
booktitle={Conference on Machine Learning and Systems},
year={2025}
}Can LLMs Design Real Hardware? A New Benchmark for RTL Design and Verification Tasks.
arXiv preprint, 2025.
@misc{pinckney2025canllms,
title={Can LLMs Design Real Hardware? A New Benchmark for RTL Design and Verification Tasks},
author={Nathaniel Pinckney and Chenhui Deng and Chia-Tung Ho and Yun-Da Tsai and Mingjie Liu and Wenfei Zhou and Brucek Khailany and Haoxing Ren},
howpublished={arXiv preprint},
year={2025}
}CirSTAG: Circuit Stability Analysis on Graph-based Manifolds.
ACM/IEEE Design Automation Conference (DAC), 2025. Best Paper Nominee
@inproceedings{cheng2025cirstag,
title={CirSTAG: Circuit Stability Analysis on Graph-based Manifolds},
author={Wuxinlin Cheng and Yihang Yuan and Chenhui Deng and Ali Aghdaei and Zhiru Zhang and Zhuo Feng},
booktitle={ACM/IEEE Design Automation Conference},
year={2025}
}ReVEAL: GNN-Guided Reverse Engineering for Formal Verification of Optimized Multipliers.
arXiv preprint, 2025.
@misc{chen2025reveal,
title={ReVEAL: GNN-Guided Reverse Engineering for Formal Verification of Optimized Multipliers},
author={Chen Chen and Daniela Kaufmann and Chenhui Deng and Zhan Song and Hongce Zhang and Cunxi Yu},
howpublished={arXiv preprint},
year={2025}
}SAGMAN: Stability Analysis of Graph Neural Networks on the Manifolds.
arXiv preprint, 2024.
@misc{cheng2024sagman,
title={SAGMAN: Stability Analysis of Graph Neural Networks on the Manifolds},
author={Wuxinlin Cheng and Chenhui Deng and Ali Aghdaei and Zhiru Zhang and Zhuo Feng},
howpublished={arXiv preprint},
year={2024}
}Vesper: A Versatile Sparse Linear Algebra Accelerator With Configurable Compute Patterns.
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD), 2024.
@article{jin2024vesper,
title={Vesper: A Versatile Sparse Linear Algebra Accelerator With Configurable Compute Patterns},
author={Hanchen Jin and Zichao Yue and Zhongyuan Zhao and Yixiao Du and Chenhui Deng and Nitish Srivastava and Zhiru Zhang},
journal={IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems},
year={2024}
}Special Session: Machine Learning for Embedded System Design.
International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS), 2023.
@inproceedings{alcorta2023special,
title={Special Session: Machine Learning for Embedded System Design},
author={Erika S. Alcorta Lozano and Andreas Gerstlauer and Chenhui Deng and Qi Sun and Zhiru Zhang and Callie Xu and Lisa Wu Wills and others},
booktitle={International Conference on Hardware/Software Codesign and System Synthesis},
year={2023}
}Machine Learning for Agile FPGA Design.
Machine Learning Applications in Electronic Design Automation, Springer, 2022. Book Chapter
@incollection{pal2022machine,
title={Machine Learning for Agile FPGA Design},
author={Debjit Pal and Chenhui Deng and Ecenur Ustun and Cunxi Yu and Zhiru Zhang},
booktitle={Machine Learning Applications in Electronic Design Automation},
year={2022}
}Layout Symmetry Annotation for Analog Circuits with Graph Neural Networks.
Asia and South Pacific Design Automation Conference (ASP-DAC), 2021.
@inproceedings{gao2021layout,
title={Layout Symmetry Annotation for Analog Circuits with Graph Neural Networks},
author={Xiaohan Gao and Chenhui Deng and Mingjie Liu and Zhiru Zhang and David Z. Pan and Yibo Lin},
booktitle={Asia and South Pacific Design Automation Conference},
year={2021}
}GLAIVE: Graph Learning Assisted Instruction Vulnerability Estimation.
Design, Automation and Test in Europe (DATE), 2021.
@inproceedings{jiao2021glaive,
title={GLAIVE: Graph Learning Assisted Instruction Vulnerability Estimation},
author={Jiajia Jiao and Debjit Pal and Chenhui Deng and Zhiru Zhang},
booktitle={Design, Automation and Test in Europe Conference},
year={2021}
}Accurate and Efficient Representation Learning on Large-Scale Graphs.
Ph.D. Dissertation, Cornell University, 2024.
@phdthesis{deng2024thesis,
title={Accurate and Efficient Representation Learning on Large-Scale Graphs},
author={Chenhui Deng},
school={Cornell University},
year={2024}
}Earned my PhD after completing research on large-scale graph learning and circuit applications.
Presented recent research progress at CDSC, UCLA.
Received the 2022 Qualcomm Innovation Fellowship as one of 19 winning teams selected across North America.