Omin Kwon

Hi, I'm Omin Kwon, an ML systems researcher at Seoul National University.

Looking for PhD positions starting in Fall 2027.

About

I am an M.S. candidate in the Department of Computer Science and Engineering at Seoul National University. I am fortunate to be advised by Prof. Jae W. Lee, Director of the AI Institute of Seoul National University, and am a member of the Architecture and Code Optimization (ARC) Lab. I also collaborate with Prof. Sehoon Kim, an Assistant Professor at KAIST AI, who co-advises my current research in ML systems.

I am interested in efficient computing through the co-design of algorithms, systems, and architectures. My current research focuses on improving the inference efficiency of diffusion language models and hybrid-attention models through ML algorithm-system co-design. In the long term, I aim to build reliable and efficient long-horizon AI and robotic agents.

Research

ML Systems

Building software systems to improve hardware utilization and serving throughput for emerging AI models.

ML Algorithms

Designing algorithms for sparse attention and quantization to improve the efficiency of foundation models.

Computer Architecture

Exploring memory, storage, and heterogeneous systems for emerging AI workloads.

Publications

  1. NeurIPS LCFM
    img_sketchssm.png
    SketchSSM: Write to the Full State, Read from a Compact Sketch
    Omin Kwon, JoongWon Shin, Minseo Kim, Kurt Keutzer, Sehoon Kim*, and Jae W. Lee*
    NeurIPS 2026 LCFM Workshop; extended version under review
  2. Under Review
    img_herald.png
    HERALD: High Throughput Block Diffusion LLM Serving via CPU-GPU Cooperative KV Cache Retrieval
    Omin Kwon, Doyeon Kim, Jongseok Park, Seung Yul Lee, Ion Stoica, and Jae W. Lee
    2026
  3. NeurIPS
    img_mage.jpg
    MAGE: All-[MASK] Block Already Knows Where to Look in Block Diffusion LLM
    Omin Kwon, Yeonjae Kim, Doyeon Kim, Minseo Kim, Yeonhong Park, and Jae W. Lee
    NeurIPS 2026& ICML 2026 AdaptFM Workshop (Best Paper Award)
  4. ICCAD
    img_star.jpg
    STAR: Improving Lifetime and Performance of High-Capacity Modern SSDs Using State-Aware Randomizer
    Omin Kwon*, Kyungjun Oh*, Jaeyong Lee, Myungsuk Kim, and Jihong Kim
    ICCAD 2025
    U.S. Patent Application with SK hynix (No. 19/569899; filed March 17, 2026)
  5. NeurIPS
    img_nestedfp.png
    NestedFP: High-Performance, Memory-Efficient Dual-Precision Floating Point Support for LLMs
    Haeun Lee, Omin Kwon, Yeonhong Park, and Jae W. Lee
    NeurIPS 2025
  6. CAL
    img_aide.png
    AiDE: Attention-FFN Disaggregated Execution for Cost-Effective LLM Decoding on CXL-PNM
    KyungSoo Kim, Omin Kwon, Yeonhong Park, and Jae W. Lee
    IEEE CAL, 2025

News

Sep 30, 2026 Our paper SketchSSM has been accepted to the NeurIPS 2026 Long Context Foundation Models (LCFM) Workshop!
Sep 25, 2026 Our paper MAGE has been accepted to NeurIPS 2026!
Jul 17, 2026 Started a research collaboration with Prof. Sehoon Kim on improving the inference efficiency of emerging hybrid-attention models that incorporate linear attention.
Jul 12, 2026 Our paper MAGE received the Best Paper Award at the ICML 2026 AdaptFM Workshop among 123 accepted papers from 218 submissions!
Jun 27, 2026 Our paper MAGE has been accepted as an oral paper to the ICML 2026 AdaptFM Workshop (1.8%)!
Mar 01, 2026 Visited the FAST Lab at the University of Illinois Urbana-Champaign (UIUC), hosted by Prof. Nam Sung Kim, a leading researcher in computer architecture.
Sep 18, 2025 Our paper NestedFP has been accepted to NeurIPS 2025!
Sep 01, 2025 Graduated from SNU ECE with a B.S. and began my M.S. studies in SNU CSE, joining ARC Lab, where I am fortunate to be advised by Prof. Jae W. Lee.
Aug 07, 2025 Our paper STAR has been accepted to ICCAD 2025! Special thanks to Prof. Jihong Kim and Prof. Myungsuk Kim for giving an undergraduate student this opportunity. A patent for this SSD technology is being prepared in collaboration with SK Hynix.
Aug 05, 2025 Our paper AiDE has been accepted to IEEE Computer Architecture Letters (CAL)!
Jan 01, 2025 Won the Grand Prize (Minister of Science and ICT Award) at the 2024 AI Semiconductor Idea and Design Contest for Resolution Aware Diffusion Accelerator Leveraging Computational Staleness. [Edaily]
Dec 01, 2024 Won the Grand Prize (Deputy Prime Minister & Minister of Education Award) at the 2024 STOB League Semiconductor Contest. [Dong-A Ilbo]
Dec 01, 2024 Joined ARC Lab as a research intern.
Aug 01, 2024 Ranked 1st in the Introduction to Operating Systems course at SNU CSE.
Dec 01, 2023 Ranked 1st in the Introduction to Computer Networking course at SNU ECE.
Aug 01, 2023 Ranked 1st in both the Introduction to Machine Learning and Computer Architecture courses at SNU ECE.
Feb 01, 2023 Completed my service in the Republic of Korea Air Force and returned to civilian life—more disciplined, slightly wiser, and more energized than ever.

Experience & Education

Full CV

Experience

2026-09 — present

Teaching Assistant, Advanced Computer Architecture (4190.571)

Seoul National University · Seoul, South Korea

2026-03 — 2026-05

Visiting Researcher

University of Illinois Urbana-Champaign (UIUC) · Urbana, IL, USA

2025-03 — 2025-06

Microelectronics (430.207B) Tutor

Seoul National University · Seoul, South Korea

2024-09 — 2024-12

Digital System Design (430.315) Tutor

Seoul National University · Seoul, South Korea

2024-03 — 2024-06

Computer Organization and Design (430.322) Tutor

Seoul National University · Seoul, South Korea

2021-05 — 2023-02

Military Service

Republic of Korea Air Force · South Korea

Education

2025-09 — present

M.S. in Computer Science and Engineering

Seoul National University · Seoul, South Korea

2019-03 — 2025-08

B.S. in Electrical and Computer Engineering (summa cum laude)

Seoul National University · Seoul, South Korea

2017-03 — 2019-03

High School Diploma in Science

Hansung Science High School · Seoul, South Korea

Early graduation with academic excellence

Contact

For research discussions and collaborations, email me at om0127@snu.ac.kr.