publications

2026

1. Multi-Chiplet GPUs

   Making Locality-aware GEMM Compatible with Page-Granularity Placement on Chiplet GPUs

   Euijun Chung, Jae Hyung Ju, and Hyesoon Kim

   IEEE Computer Architecture Letters, 2026

   Bib PDF

   @article{chung2026making,
     title = {Making Locality-aware GEMM Compatible with Page-Granularity Placement on Chiplet GPUs},
     author = {Chung, Euijun and Ju, Jae Hyung and Kim, Hyesoon},
     journal = {IEEE Computer Architecture Letters},
     year = {2026},
     publisher = {IEEE},
   }

2. LLM Inference

   Characterizing CPU-Induced Slowdowns in Multi-GPU LLM Inference

   Euijun Chung, Yuxiao Jia, Aaron Jezghani, and 1 more author

   In 2026 IEEE international symposium on workload characterization (IISWC), 2026

   Best Paper Candidate

   arXiv Bib PDF IEEE Spectrum

   @inproceedings{chung2026characterizing,
     title = {Characterizing CPU-Induced Slowdowns in Multi-GPU LLM Inference},
     author = {Chung, Euijun and Jia, Yuxiao and Jezghani, Aaron and Kim, Hyesoon},
     booktitle = {2026 IEEE international symposium on workload characterization (IISWC)},
     year = {2026},
     note = {Best Paper Candidate},
     organization = {IEEE},
   }

3. GPU Memory System

   TileLens: Efficiently Using Large-Granularity Memory Systems with Transparent Two-Dimensional Memory Layout

   Jae Hyung Ju^*, Euijun Chung^*, Hritvik Taneja, and 4 more authors

   2026

   Preprint

   arXiv Bib

   @misc{ju2026tilelens,
     title = {TileLens: Efficiently Using Large-Granularity Memory Systems with Transparent Two-Dimensional Memory Layout},
     author = {Ju, Jae Hyung and Chung, Euijun and Taneja, Hritvik and Saxena, Anish and Jeong, Shinnung and Kim, Hyesoon and Qureshi, Moinuddin K},
     note = {Preprint},
     year = {2026},
     eprint = {2607.04031},
     archiveprefix = {arXiv},
     primaryclass = {cs.AR},
   }

4. Multi-Chiplet GPUs

   A Fast Locality Simulator for GEMM Design-Space Exploration on Multi-Chiplet GPUs

   Euijun Chung and Hyesoon Kim

   2026

   Poster presentation at the Workshop on Modeling & Simulation of Systems and Applications (MODSIM 2026)

   arXiv Bib PDF Code Slides

   @misc{chung2026fast,
     title = {A Fast Locality Simulator for GEMM Design-Space Exploration on Multi-Chiplet GPUs},
     author = {Chung, Euijun and Kim, Hyesoon},
     note = {Poster presentation at the Workshop on Modeling & Simulation of Systems and Applications (MODSIM 2026)},
     year = {2026},
     eprint = {2606.11716},
     archiveprefix = {arXiv},
     primaryclass = {cs.AR},
   }

5. GPU Fault Injection

   TensorDynamic: Bridging Application- and Instruction-Level Fault Injection for DNN Tensor Core Execution

   Yuxiao Jia, Euijun Chung, Huanzhi Pu, and 2 more authors

   In 2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), 2026

   Bib PDF Slides

   @inproceedings{jia2026tensordynamic,
     title = {TensorDynamic: Bridging Application- and Instruction-Level Fault Injection for DNN Tensor Core Execution},
     author = {Jia, Yuxiao and Chung, Euijun and Pu, Huanzhi and Feinberg, Ben and Kim, Hyesoon},
     booktitle = {2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)},
     year = {2026},
     artifact_available = {https://github.com/yuxiaojia/TensorDynamic},
     artifact_functional = {https://github.com/yuxiaojia/TensorDynamic},
     artifact_reproduced = {https://github.com/yuxiaojia/TensorDynamic}
   }

6. GPU Simulation

   Macsim Mini: A Lightweight Cycle-Level GPU Simulator for Architecture Education

   Euijun Chung^*, Saurabh Singh^*, Huanzhi Pu, and 5 more authors

   In 2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), 2026

   Bib PDF Slides

   @inproceedings{chung2026macsim,
     title = {Macsim Mini: A Lightweight Cycle-Level GPU Simulator for Architecture Education},
     author = {Chung, Euijun and Singh, Saurabh and Pu, Huanzhi and Jia, Yuxiao and Kar, Anurag and Jijina, Sam and Madeira, Scott and Kim, Hyesoon},
     booktitle = {2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)},
     year = {2026},
   }

2025

1. GPU Simulation

   Swift and Trustworthy Large-Scale GPU Simulation with Fine-Grained Error Modeling and Hierarchical Clustering

   Euijun Chung, Seonjin Na, Sung Ha Kang, and 1 more author

   In Proceedings of the 58th IEEE/ACM International Symposium on Microarchitecture, 2025

   Bib PDF Slides

   @inproceedings{chung2025swift,
     title = {Swift and Trustworthy Large-Scale GPU Simulation with Fine-Grained Error Modeling and Hierarchical Clustering},
     author = {Chung, Euijun and Na, Seonjin and Kang, Sung Ha and Kim, Hyesoon},
     booktitle = {Proceedings of the 58th IEEE/ACM International Symposium on Microarchitecture},
     pages = {1397--1411},
     year = {2025},
     artifact_available = {https://github.com/ejchung0406/STEM-AE},
     artifact_functional = {https://github.com/ejchung0406/STEM-AE}
   }

2. GPU Memory Safety

   Let-Me-In:(Still) Employing In-pointer Bounds Metadata for Fine-grained GPU Memory Safety

   Jaewon Lee, Euijun Chung, Saurabh Singh, and 4 more authors

   In 2025 IEEE International Symposium on High Performance Computer Architecture (HPCA), 2025

   Bib PDF Slides

   @inproceedings{lee2025let,
     title = {Let-Me-In:(Still) Employing In-pointer Bounds Metadata for Fine-grained GPU Memory Safety},
     author = {Lee, Jaewon and Chung, Euijun and Singh, Saurabh and Na, Seonjin and Kim, Yonghae and Lee, Jaekyu and Kim, Hyesoon},
     booktitle = {2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)},
     pages = {1648--1661},
     year = {2025},
     organization = {IEEE},
   }

3. GPU-SSD Arch.

   Contention-Aware GPU Thread Block Scheduler for Efficient GPU-SSD

   Xueyang Liu, Seonjin Na, Euijun Chung, and 3 more authors

   IEEE Computer Architecture Letters, 2025

   Bib PDF

   @article{liu2025contention,
     title = {Contention-Aware GPU Thread Block Scheduler for Efficient GPU-SSD},
     author = {Liu, Xueyang and Na, Seonjin and Chung, Euijun and Cao, Jiashen and Yang, Jing and Kim, Hyesoon},
     journal = {IEEE Computer Architecture Letters},
     year = {2025},
     publisher = {IEEE},
   }

4. Open-source GPU

   Hardware vs. Software Implementation of Warp-Level Features in Vortex RISC-V GPU

   Huanzhi Pu, Rishabh Ravi, Shinnung Jeong, and 5 more authors

   In OSSMPIC - Open Source Solutions for Massively Parallel Integrated Circuits (Co-located with DATE 2025), 2025

   Bib PDF

   @inproceedings{pu2025hardware,
     title = {Hardware vs. Software Implementation of Warp-Level Features in Vortex RISC-V GPU},
     author = {Pu, Huanzhi and Ravi, Rishabh and Jeong, Shinnung and Subramanya, Udit and Chung, Euijun and Zhao, Jisheng and Ahn, Chihyo and Kim, Hyesoon},
     booktitle = {OSSMPIC - Open Source Solutions for Massively Parallel Integrated Circuits (Co-located with DATE 2025)},
     year = {2025},
   }

2024

1. GPU Simulation

   Allegro: GPU Simulation Acceleration for Machine Learning Workloads

   Euijun Chung, Seonjin Na, and Hyesoon Kim

   In Machine Learning for Computer Architecture and Systems 2024 (Co-located with ISCA 2024), 2024

   Bib PDF Slides

   @inproceedings{chung2024allegro,
     title = {Allegro: GPU Simulation Acceleration for Machine Learning Workloads},
     author = {Chung, Euijun and Na, Seonjin and Kim, Hyesoon},
     booktitle = {Machine Learning for Computer Architecture and Systems 2024 (Co-located with ISCA 2024)},
     year = {2024},
   }

2016

1. Plasma Physics

   Experimental analysis on the 0 dimensional plasma model in an inductively coupled plasma (ICP)

   M-J Lee and Euijun Chung

   New Physics: Sae Mulli, 2016

   Bib PDF

   @article{lee2016experimental,
     title = {Experimental analysis on the 0 dimensional plasma model in an inductively coupled plasma (ICP)},
     author = {Lee, M-J and Chung, Euijun},
     journal = {New Physics: Sae Mulli},
     volume = {66},
     number = {9},
     pages = {1183--1189},
     year = {2016},
     publisher = {The Korean Physical Society},
   }