Jae-Won Chung

Ph.D. Candidate @ UMich CSE

Summary

I'm a fifth year PhD candidate in CSE at the University of Michigan. I build efficient software systems for deep learning, with a focus on the efficient management of not only time, but also energy.

I view power and energy as fundamental systems resources that are worth carefully optimizing and allocating, not only in hardware, but also from software. Doing so provides automatic downstream benefits, such as reducing operational expenses, alleviating power delivery pressure for datacenters, and allowing the hardware to truly max out on performance.

I lead the ML.ENERGY initiative as part of my research and open-source efforts. I am fortunate to be advised by Professor Mosharaf Chowdhury and be part of SymbioticLab.

Selected Publications

Cornserve: A Distributed Serving System for Any-to-Any Multimodal Models

Preprint, 2026

Jae-Won Chung*, Jeff J. Ma*, Jisang Ahn, Yizhuo Liang, Akshay Jajoo, Myungjin Lee, Mosharaf Chowdhury (* Equal Contribution)

GPU-to-Grid: Voltage Regulation via GPU Utilization Control

Preprint, 2026

Zhirui Liang, Jae-Won Chung, Mosharaf Chowdhury, Jiasi Chen, Vladimir Dvorkin

Where Do the Joules Go? Diagnosing Inference Energy Consumption

Preprint, 2026

Jae-Won Chung, Ruofan Wu, Jeff J. Ma, Mosharaf Chowdhury

Kareus: Joint Reduction of Dynamic and Static Energy in Large Model Training

Preprint, 2026

Ruofan Wu, Jae-Won Chung, Mosharaf Chowdhury

Cornserve: Efficiently Serving Any-to-Any Multimodal Models

Preprint, 2025

Jeff J. Ma*, Jae-Won Chung*, Jisang Ahn, Yizhuo Liang, Akshay Jajoo, Myungjin Lee, Mosharaf Chowdhury (* Equal Contribution)

The ML.ENERGY Benchmark: Toward Automated Inference Energy Measurement and Optimization

NeurIPS D&B spotlight, 2025 (spotlight acceptance rate = 2.81%)

Jae-Won Chung, Jeff J. Ma, Ruofan Wu, Jiachen Liu, Oh Jun Kweon, Yuxuan Xia, Zhiyu Wu, Mosharaf Chowdhury

Perseus: Reducing Energy Bloat in Large Model Training

SOSP, 2024 (Acceptance rate = 17.34%)

Jae-Won Chung, Yile Gu, Insu Jang, Luoxi Meng, Nikhil Bansal, Mosharaf Chowdhury

Andes: Defining and Enhancing Quality-of-Experience in LLM-Based Text Streaming Services

Preprint, 2024

Jiachen Liu, Jae-Won Chung, Zhiyu Wu, Fan Lai, Myungjin Lee, Mosharaf Chowdhury

Zeus: Understanding and Optimizing GPU Energy Consumption of DNN Training

USENIX NSDI, 2023 (Acceptance rate = 18.38%)

Jie You*, Jae-Won Chung*, Mosharaf Chowdhury (* Equal Contribution)

Selected Talks

Energy and Power as First-Class ML Design Metrics

UW-Madison madSystems seminar | October 2025

  • Energy
  • MLSys
How to Create Mediocre Open-Source Repositories

University of Michigan (SymbioticLab lunch seminar) | August 2024

Energy-Efficient Software Systems for Machine Learning

Seoul National University | October 2023

  • Energy
  • MLSys
Energy-Efficient Deep Learning with PyTorch and Zeus

PyTorch Conference | October 2023

Energy-Efficient Deep Learning with Zeus

Massachusetts Institute of Technology | September 2023

  • Energy
  • MLSys
Memory Plus Meta-Learning

Deepest | August 2019

Open-Source Projects

Number of stars and forks are as of April 8th, 2026.
Cornserve (122 8)

An efficient inference platform for any-to-any multimodal models.

OpenG2G (9 0)

A Modular simulation library for AI datacenter-grid interaction.

BERT4Rec-VAE-Pytorch (413 98)

Implementation of BERT4Rec and Netflix VAE recommendation models.

  • Python
  • PyTorch
  • RecSys
  • |
  • GitHub
Reason (197 4)

A shell for research papers. Supports UNIX-like commands that instead work on a set of research papers.

Pegasus (33 3)

An SSH command runner with a focus on simplicity. Useful when you have a bunch of commands to run and a bunch of SSH nodes available.

Education

  • PhD, Computer Science and Engineering
    University of Michigan
    Sep 2021 - May 2027 (expected)
  • MS, Computer Science and Engineering
    University of Michigan
    Sep 2021 - Apr 2023
  • BS, Electrical and Computer Engineering
    Summa cum laude
    Seoul National University, South Korea
    Mar 2015 - Aug 2021

Honors & Awards

  • Rackham Predoctoral Fellowship
    Full tuition and stipend for three terms.
    April 2026
  • Laude Institute Slingshot Grant
    $25k USD for ML.ENERGY.
    Dec 2025
  • GitHub Secure Open Source Fund
    $10k USD for Zeus.
    Aug 2025
  • Mozilla Technology Fund
    $50k USD for Zeus.
    Feb 2024
  • Salesforce Research Award
    $20k USD for ML.ENERGY.
    Jan 2024

Teaching

  • CSE585: Systems for GenAI
    Graduate Student Instructor. Three lectures on GenAI and Systems for GenAI fundamentals.
    Fall 2025
  • Undergrad Operating Systems
    Lead TA. Linux kernel lectures, four Linux-based term projects, and team design reviews.
    Spring 2021
  • Undergrad Computer Architecture
    Peer tutor. Gave 30 hours of online lecture. Best tutor award!
    Fall 2020

Community Service