on-going R&D history

Software Engineer (Technical Research Personnel for Military Service, 병역복무), Deep Learning Compiler, DEEPX Co., Ltd., Republic of Korea (2021.10 - 2025.02)

Doctor of Philosophy (Ph.D.), Computer Science and Engineering (CSE), Sogang University, Republic of Korea (2019.09 - 2021.08, 2025.03 - Present)

Master of Engineering (M.S.), Computer Science and Engineering (CSE), Sogang University, Republic of Korea (2017.09 - 2019.08)

Exchange Visiting Scholar, Electrical and Computer Engineering (ECE), The University of Texas at San Antonio (UTSA), TX, USA (2018.07 - 20218.08)

Bachelor of Engineering (B.S.), Computer Science and Engineering (CSE), Sogang University, Republic of Korea (2013.03 - 2017.08)

[@DEEPX] (2021.10 ~ 2025.02)

• Deep Learning Compiler Engineer (My mission was to build a backend compilation considering both deep learning architecture and NPU hardware architecture, aiming to maximize the potential hardware performance during deep learning model operations.)

  • Designed and implemented a multi-level intermediate representation (MLIR)-based memory operation processing for backend code generation (keyword: MLIR).

  • Designed and implemented layer fusion techniques into backend code generation (keyword: Layer Fusion).

  • Designed and implemented a simultaneous memory bank access pattern-aware tensor placement strategy in backend code generation (keyword: Variable Memory).

  • Supported instruction generation for operating vision transformers on the DEEPX neural processing unit (keyword: Vision Transformer Architecture).

  • Designed and implemented advanced data reuse optimization algorithms in backend code generation (keyword: NPU Memory Hierarchy).

  • Invented two U.S. patents (currently under review)

[@Sogang University] (2017.09 ~ 2021.08, 2025.03 ~ Present)

• [Ph.D.] Graduate Research Assistant (My mission was to design and implement principles for a storage engine that supports managing and processing unstructured/semi-structured data.)

  • Designed and implemented performance and namespace isolation and secondary indexing capability in key-value solid-state drives (keyword: KV-Store).

  • Invented a U.S. patent

• [M.S.] Graduate Research Assistant (My mission was to design and implement machine learning-based secure device solutions against unauthorized access.)

  • Designed and implemented the machine learning-based detection and backup in solid-state drives against ransomware attacks (keywords: Ransomware and SSD).

  • Implemented access control for non-white list users to files in FUSE filesystem (keyword: Filesystem).

[International]

1. [arXiv] Junyoung Kim, Youngrok Kim, Siyeol Jung, Donghyun Min, Crafting Query-Aware Selective Attention For Single Image Super-resolution, arXiv:2504.06634, 2025. [Paper]

2. [Best Paper Nominee] [ACM ICS] Juhyeon Lee, Insung Bahk, Hoseung Kim, Sinjin Jeong, Suyeon Lee, Donghyun Min (corresponding author), An Autonomous Parallelization Of Transformer Model Inference On Heterogeneous Edge Devices, The ACM International Conference on Supercomputing (ACM ICS), 2024. (overall acceptance rate: 28%) [Paper] [Univ News] (It was my first R&D project as the research director.)

3. [ACM Systor] Seungjin Lee, Chang-Gyu Lee, Donghyun Min, Inhyuk Park, Woosuk Chung, Anand Sivasubramaniam, Youngjae Kim, Iterator Interface Extended LSM-tree-based KVSSD for Range Queries, The ACM International Systems and Storage Conference (ACM SYSTOR), 2023. (acceptance rate: 40%) [Paper]

4. [ACM TECS] Donghyun Min, Kihyun Kim, Chaewon Moon, Awais Khan, Seungjin Lee, Changhwan Yun, Woosuk Chung, Youngjae Kim, A Multi-Tenant Key-Value SSD with Secondary Index for Search Query Processing and Analysis, ACM Transactions on Embedded Computing Systems (ACM TECS), 2023. [Paper]

5. [IEEE TCAD] Donghyun Min, Yungwoo Ko, Ryan Walker, Junhee Lee, Youngjae Kim, A Content-based Ransomware Detection and Backup Solid-State Drive for Ransomware Defense, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (IEEE TCAD), 2022. [Paper]

6. [Honorable Mention] [ACM HotStorage] Donghyun Min, Youngjae Kim, Isolating Namespace and Performance in Key-Value SSDs for Multi-tenant Environments, (Regular) ACM Workshop on Hot Topics in Storage and File Systems (ACM HotStorage), 2021. (acceptance rate: 37.5%) [Paper]

7. [ACM ASIACCS] Jinwoo Ahn, Junghee Lee, Yungwoo Ko, Donghyun Min, Jiyun Park, Sungyong Park, Youngjae Kim, Diskshield: A Data Tamper-resistant Storage for Intel SGX, ACM ASIA Conference on Computer and Communications Security (ACM ASIACCS), 2020. [Paper]

8. [arXiv] Jinwoo Ahn, Donggyu Park, Chang-gyu Lee, Donghyun Min, Junghee Lee, Sungyong Park, Qian Chen, Youngjae Kim, KEY-SSD: Access-Control Drive to Protect Files from Ransomware Attacks, arXiv:1904.05012, 2019. [Paper]

9. [Best Paper] [IEEE CAL] Donghyun Min, Donggyu Park, Jinwoo Ahn, Ryan Walker, Junghee Lee, Sungyong Park, Youngjae Kim, Amoeba: An Autonomous Backup and Recovery SSD for Ransomware Defense, IEEE Computer Architecture Letter (IEEE CAL), 2018. (acceptance rate: approximately 20%) [Paper][Univ News] (It was my first R&D project as the lead author.)

10. [USENIX FAST] Junghee Lee, Donghyun Min, Donggyu Park, Sungyong Park, Youngjae Kim, Low-Cost Autonomous Backup Mechanism in SSD against Ransomware, (WiP) USENIX Conference on File and Storage Technologies (USENIX FAST), 2018. [Paper]

[Domestic]

1. [KIISE] Donghyun Min, Youngjae Kim, A Study of Log-Structured Merge-tree Based Key-Value SSDs, 2021. [Paper]

2. [KSC] Jiyun Park, Donghyun Min, Junghee Lee, Youngjae Kim, Reducing Foreground I/O Latency via Preemptive Ransomware Detection on Ransomware Attack Tolerant SSD, Korea Software Congress (KSC), 2020.

3. [KSC] Jeewon Chae, Jinwoo Ahn, Donghyun Min, Junghee Lee, Youngjae Kim, TC-SSD: Time Capsule SSD for File Version Control, Korea Software Congress (KSC), 2019.

4. [KSC] Donghyun Min, Jinwoo Ahn, Youngjae Kim, Infected Data Segregation and Fast Garbage Collection on Ransomware Defense SSD, Korea Software Congress (KSC), 2018.

[Awards & Honors]

• Best Paper Award Nominee, ACM ICS, 2024.06

• Honorable Mention, ACM HotStorage, 2021.07

• Best Paper Award, IEEE CAL, 2018.11

• Academic Improvement Scholarship (A) Award, Sogang University, 2015.08

• National Science and Engineering Advancement Scholarship Qualification, The Republic of South Korea, 2012.02

[Invited Writing & Talks]

• Invited journal contribution to KIISE, 2021.07

• Invited paper session in KCC, 2019.07

• Best of CAL session in IEEE HPCA, 2019.02