Sangjin Choi

Sangjin Choi

Ph.D., School of Computing, KAIST

I am a Ph.D. graduate advised by Youngjin Kwon.
My research focuses on building efficient ML systems.

MoE Serving PD Disaggregation Energy-Efficient ML GPU Memory Management
Email / CV / Google Scholar / CASYS Lab

News

Education

Experience

Publications

MLSys '26
BEAM: Joint Resource–Power Optimization for Energy-Efficient LLM Inference under SLO Constraints
Hyunjae Lee, Sangjin Choi, Seungjae Lim, Youngjin Kwon
Paper
EuroSys '26
MTTM: Dynamic Fast Memory Partitioning with Bandwidth Optimization for Multi-tenant Cloud
Changjun Lee, Sangjin Choi, Youngjin Kwon
Paper
Findings of NAACL '25
Lossless Acceleration of Large Language Models with Hierarchical Drafting based on Temporal Locality in Speculative Decoding
Sukmin Cho, Sangjin Choi, Taeho Hwang, Jeongyeon Seo, Soyeong Jeong, Huije Lee, Hoyun Song, Jong C. Park, Youngjin Kwon
Paper
ICCAD '23
PRIMO: A Full-Stack Processing-in-DRAM Emulation Framework for Machine Learning Workloads
Jaehoon Heo, Yongwon Shin, Sangjin Choi, Sungwoong Yune, Jung-Hoon Kim, Hyojin Sung, Youngjin Kwon, Joo-Young Kim
Paper
USENIX ATC '23
EnvPipe: Performance-preserving DNN Training Framework for Saving Energy
Sangjin Choi, Inhoe Koo, Jeongseob Ahn, Myeongjae Jeon, Youngjin Kwon
Paper
USENIX ATC '22
Memory Harvesting in Multi-GPU Systems with Hierarchical Unified Virtual Memory
Sangjin Choi*, Taeksoo Kim*, Jinwoo Jeong, Rachata Ausavarungniurn, Myeongjae Jeon, Youngjin Kwon, Jeongseob Ahn (*equal contribution)
Paper