Bodun Hu

I am a software engineer at Google, where I work on improving the reliability of Gemini training. My research focuses on high-performance systems for machine learning, particularly LLM training and serving, with interests spanning operating systems, networking, and compute-communication co-design.

I earned my Ph.D. in Computer Science from UT Austin, advised by Aditya Akella in the UTNS lab. I also received my M.S. and B.S. in Computer Science from UT Austin. Before my Ph.D., I worked with Christopher Rossbach in the SCEA lab on GPU benchmarking and methods for enabling OS kernels to access accelerators.

I previously interned at Intel with Theo Jepsen and Georgios Nikolaidis, and at Meta's AI and Systems Co-design team, where I developed efficient communication collectives for LLM training under the supervision of Ching-Hsiang Chu.

I occasionally share notes and write-ups on my blog.

Contact: bodunhu at gmail.com
PGP key: 76F092F8CD673517

Publications

• CUCo: An Agentic Framework for Compute and Communication Co-design, Preprint
  Bodun Hu*, Yoga Sri Varshan V*, Saurabh Agarwal, Aditya Akella
• CARE-RFT: Confidence-Anchored Reinforcement Finetuning for Reliable Reasoning in Large Language Models, Preprint
  Shuozhe Li, Jincheng Cao, Bodun Hu, Aryan Mokhtari, Leqi Liu, Amy Zhang
• SYMPHONY: Enabling Compute-Memory Disaggregation in LLM Serving Systems, NSDI 2026
  Saurabh Agarwal, Bodun Hu, Anyong Mao, Aditya Akella, Shivaram Venkataraman
• ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models, NeurIPS 2025
  Liyan Tang, Grace Kim, Xinyu Zhao, Thom Lake, Wenxuan Ding, Fangcong Yin, Prasann Singhal, Manya Wadhwa, Zeyu Leo Liu, Zayne Sprague, Ramya Namuduri, Bodun Hu, Juan Diego Rodriguez, Puyuan Peng, Greg Durrett
• Patchwork: A Unified Framework for RAG Serving, ArXiv
  Bodun Hu*, Luis Pabon*, Saurabh Agarwal, Aditya Akella
• StitchLLM: Serving LLMs, One Block at a Time, ACL 2025
  Bodun Hu, Shuozhe Li, Saurabh Agarwal, Myungjin Lee, Akshay Jajoo, Jiamin Li, Le Xu, Geon-Woo Kim, Donghyun Kim, Hong Xu, Amy Zhang, Aditya Akella
• MOSEL: Inference Serving Using Dynamic Modality Selection, EMNLP 2024
  Bodun Hu, Le Xu, Jeongyoon Moon, Neeraja J. Yadwadkar, Aditya Akella
• FFN-SkipLLM: A Hidden Gem for Autoregressive Decoding with Adaptive Feed Forward Skipping, EMNLP 2024
  Ajay Jaiswal, Bodun Hu, Lu Yin, Yeonju Ro, Shiwei Liu, Tianlong Chen, Aditya Akella
• Towards a Machine Learning-Assisted Kernel with LAKE, ASPLOS 2023
  Henrique Fingler, Isha Tarte, Hangchen Yu, Ariel Szekely, Bodun Hu, Aditya Akella, Christopher J. Rossbach
• Altis: Modernizing GPGPU Benchmarks, ISPASS 2020
  Bodun Hu, Christopher J. Rossbach

Teaching

• Fall 2025: CS 395T: Advanced Topics in Systems and GenAI
• Fall 2024: CS 378 System For Machine Learning and Big Data
• Spring 2020: CS 378 Multicore Operating System Implementation

Service

Reviewer for ACL 2025, ACL 2026, ACL-SRW 2025, ICLR 2026, CAIS 2026, NeurIPS 2026, DocInsights 2026, and ICLR 2027.