Yunkai Liang

I am a Ph.D. candidate in the School of Computer Science and Engineering at Sun Yat-sen University, advised by Prof. Zhi Zhou. Before that, I received my B.Eng. degree in software engineering from Sun Yat-sen University. My research interests focus on large language model (LLM) inference and cloud computing system. During my Ph.D. program, I also gained industrial experience as a research intern at Huawei Cloud, working with Dr. Pengfei Zuo on LLM inference system.

Publications

2026

PPoPP

Laser: Unlocking Layer-Level Scheduling for Efficient Multi-SLO LLM Serving

Jianxiong Liao, Quanxing Dong, Yunkai Liang, Zhi Zhou, and Xu Chen

In Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming, 2026

Bib

@inproceedings{liao2026laser,
  title = {Laser: Unlocking Layer-Level Scheduling for Efficient Multi-SLO LLM Serving},
  author = {Liao, Jianxiong and Dong, Quanxing and Liang, Yunkai and Zhou, Zhi and Chen, Xu},
  booktitle = {Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming},
  pages = {509--521},
  year = {2026},
}

2025

arxiv

Injecting Adrenaline into LLM Serving: Boosting Resource Utilization and Throughput via Attention Disaggregation

Yunkai Liang, Zhangyu Chen, Pengfei Zuo, Zhi Zhou, Xu Chen, and Zhou Yu

arXiv preprint arXiv:2503.20552, 2025

Bib PDF Blog Code

@article{liang2025injecting,
  title = {Injecting Adrenaline into LLM Serving: Boosting Resource Utilization and Throughput via Attention Disaggregation},
  author = {Liang, Yunkai and Chen, Zhangyu and Zuo, Pengfei and Zhou, Zhi and Chen, Xu and Yu, Zhou},
  journal = {arXiv preprint arXiv:2503.20552},
  year = {2025},
}

arxiv

Prefill-decode aggregation or disaggregation? unifying both for goodput-optimized llm serving

Chao Wang, Pengfei Zuo, Zhangyu Chen, Yunkai Liang, Zhou Yu, and Ming-Chang Yang

arXiv preprint arXiv:2508.01989, 2025

Bib PDF

@article{wang2025prefill,
  title = {Prefill-decode aggregation or disaggregation? unifying both for goodput-optimized llm serving},
  author = {Wang, Chao and Zuo, Pengfei and Chen, Zhangyu and Liang, Yunkai and Yu, Zhou and Yang, Ming-Chang},
  journal = {arXiv preprint arXiv:2508.01989},
  year = {2025},
}

2024

OSDI

(Poster) PipeDecode: Efficient LLM Inference using Pipelines within Decoding

Yunkai Liang, Bin Gao, Pengfei Zuo, Zhi Zhou, and Xu Chen

In 18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24), 2024

Bib HTML PDF

@inproceedings{liang2024pipdecode,
  title = {(Poster) PipeDecode: Efficient LLM Inference using Pipelines within Decoding},
  author = {Liang, Yunkai and Gao, Bin and Zuo, Pengfei and Zhou, Zhi and Chen, Xu},
  booktitle = {18th USENIX Symposium on Operating Systems Design and Implementation (OSDI 24)},
  year = {2024},
}

2023

ICDCS

EdgeOrcher: Predictive Function Orchestration for Serverless-Based Edge Native Applications

Yunkai Liang, Zhi Zhou, and Xu Chen

In Ph.D. student symposium, 2023 IEEE 43rd International Conference on Distributed Computing Systems (ICDCS), 2023

Bib

@inproceedings{liang2023edgeorcher,
  title = {EdgeOrcher: Predictive Function Orchestration for Serverless-Based Edge Native Applications},
  author = {Liang, Yunkai and Zhou, Zhi and Chen, Xu},
  booktitle = {Ph.D. student symposium, 2023 IEEE 43rd International Conference on Distributed Computing Systems (ICDCS)},
  pages = {1--2},
  year = {2023},
  organization = {IEEE},
}

Serving

Reviewer: DPCS’24
External Reviewer: ICML’26, Eurosys’26, NSDI’25, NAS’24