I'm Kai Lu, currently a postdoctoral fellow at Huazhong University of Science and Technology (HUST). My major is storage systems and techniques. I'm in Parallel Data Storage Lab (PDSL) led by Prof. Jiguang Wan and Prof. Changsheng Xie. PDSL has long been devoted to the research of distributed storage systems, key-value storage and AI storage.

🎓 Educations

• 2018.09-2023.06, Ph.D. degree in Computer Architecture, Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, China.
• 2014.09-2018.06, B.E. degree in Computer Science and Technology, Huazhong University of Science and Technology, China.

💻 Interests

• Disaggregated Memory: Rcmp, SepHash, Scythe
• Key-value Store: Rocksdb, TiKV, Leveldb, NStore
• Distributed Storage: Ceph, PeakFS,
• Non-volatile Memory(NVM): Sprint-AEP, Geardb, WIPE
• AI for Storage: Learned-Rocksdb, Park, Replica-Placement, Auto-Tuning-Ceph
• System for AI: DI-store, DeepSpeed, vLLM, Q-Infer

🏅 Projects and Awards

• 2022-now, Ranked #1 in the IO500 10 Node Research
• 2024, the China Postdoctoral Science Foundation Funded Project
• 2024, the Hubei Provincial Postdoctoral Innovative Talent Training Project A Grade
• 2025, the Young Scientists Fund-Type C of the National Natural Science Foundation of China

📝 Publications

2025

• Kai Lu, Qiang Wei, Yier Lin, Pengyu Liu, Haipeng Wang, Jiguang Wan*, Ting Yao, Huatao Wu, Daohui Wang. Q-Infer: Towards Efficient GPU-CPU Collaborative LLM Inference via Sparsity-Aware Dynamic Scheduling. ACM Transactions on Architecture and Code Optimization (TACO 2025, CCF-A)

• Chen Ding, Sicen Li, Kai Lu*, Ting Yao, Daohui Wang, Huatao Wu, Jiguang Wan, Zhihu Tan, Changsheng Xie. DShuffle: DPU-Optimized Shuffle Framework for Large-scale Data Processing (USENIX ATC 2025, CCF-A)

• Chen Ding, Kai Lu*, QuanYi Zhang, Zekun Ye, Ting Yao, Daohui Wang, Huatao Wu, Jiguang Wan. DFlush: DPU-Offloaded Flush for Disaggregated LSM-based Key-Value Stores (SIGMOD 2025, CCF-A)

• Wei Tao, Haocheng Lu, Xiaoyang Qu*, Bin Zhang, Kai Lu*, Jiguang Wan, Jianzong Wang. MoQAE: Mixed-Precision Quantization for Long-Context LLM Inference via Mixture of Quantization-Aware Experts (ACL 2025 Main, CCF-A)

• Junjie Li, Nan Zhang, Xiaoyang Qu, Kai Lu, Guokuan Li, Jiguang Wan, Jianzong Wang. RATE-Nav: Region-Aware Termination Enhancement for Zero-shot Object Navigation with Vision-Language Models (ACL 2025 Findings, CCF-A)

• Wei Tao, Xiaoyang Qu*, Kai Lu*, Jiguang Wan, Guokuan Li, Jianzong Wang. MADLLM: Multivariate Anomaly Detection via Pre-trained LLMs (ICME 2025, CCF-B)

• Bin Zhang, Jinggang Chen, Xiaoyang Qu, Guokuan Li, Kai Lu*, Jiguang Wan, Jing Xiao, Jianzong Wang*. RUNA: Object-level Out-of-Distribution Detection via Regional Uncertainty Alignment of Multimodal Representations (AAAI 2025, CCF-A)

2024

• Yixiao Chen, Haomai Yang, Kai Lu*, Wenlve Huang, Jibin Wang, Jiguang Wan, Jian Zhou, Fei Wu, Changsheng Xie. PeakFS: An Ultra-high Performance Parallel File System via Computing-Network-Storage Co-optimization for HPC Applications (TPDS 2024, CCF-A) Code

• Kai Lu, Siqi Zhao, Haikang Shan, Qiang Wei, Guankuan Li, Jiguang Wan, Ting Yao, Huatao Wu, Daohui Wang. Scythe: A Low-latency RDMA-enabled Distributed Transaction System for Disaggregated Memory. ACM Transactions on Architecture and Code Optimization (TACO 2024, CCF-A) Code

• Zhonghua Wang, Kai Lu*, Jiguang Wan, Hong Jiang, Zeyang Zhao, Peng Xu, Biliang Lai, Guokuan Li, and Changsheng Xie. NStore: A High-Performance NUMA-Aware Key-Value Store for Hybrid Memory (TC 2024, CCF-A) Code

• Xinhao Min, Kai Lu*, Pengyu Liu, Jiguang Wan, Changsheng Xie, Daohui Wang, Ting Yao, Huatao Wu. SepHash: A Write-Optimized Hash Index on Disaggregated Memory via Separate Segment Structure (VLDB 2024, CCF-A) Code

• Yiwen Zhang, Guokuan Li*, Kai Lu*, Jiguang Wan, Ting Yao, Huatao Wu, Daohui Wang. PhatKV: Towards an Efficient Metadata Engine for KV-based File Systems on Modern SSD (MSST 2024, CCF-B)

• Chen Ding, Jian Zhou, Kai Lu, Sicen Li, Yiqin Xiong, Jiguang Wan, Ling Zhan. D2Comp: Efficient Offload of LSM-tree Compaction with Data Processing Units on Disaggregated Storage (TACO 2024, CCF-A)

• Yuanhui Zhou, Jian Zhou, Kai Lu, Ling Zhan, Peng Xu, Peng Wu, Shuning Chen, Xian Liu, Jiguang Wan. A contract-aware and cost-effective LSM Store for Cloud Storage with Low Latency Spikes (TOS 2024, CCF-A)

2023 and before

• Zhonghua Wang, Yixing Guo, Kai Lu*, Jiguang Wan, Daohui Wang, Ting Yao, Huatao Wu. Rcmp: Reconstructing RDMA-based Memory Disaggregation via CXL (TACO 2023, CCF-A) Code

• Liang Wang, Kai Lu(co-primary author), Nan Zhang, Xiaoyang Qu, Jianzong Wang, Jiguang Wan, Guokuan Li, Jing Xiao. Shoggoth: Towards Efficient Edge-Cloud Collaborative Real-Time Video Inference via Adaptive Online Learning (DAC 2023, CCF-A)

• Kai Lu, Guokuan Li, Jiguang Wan, Ruixiang Ma, Wei Zhao. ADSTS: Automatic Distributed Storage Tuning System Using Deep Reinforcement Learning (ICPP 2022, CCF-B)

• Kai Lu, Nannan Zhao, Jiguang Wan, Changhong Fei, Wei Zhao, Tongliang Deng. RLRP: High-Efficient Data Placement with Reinforcement Learning for Modern Distributed Storage Systems (IPDPS 2022, CCF-B) Code

• Kai Lu, Nannan Zhao, Jiguang Wan, Changhong Fei, Wei Zhao, and Tongliang Deng. TridentKV: A Read-Optimized LSM-tree Based KV Store via Adaptive Indexing and Space-Efficient Partitioning (TPDS, CCF-A) Code

• Zhonghua Wang, Chen Ding, Fengguang Song, Kai Lu, Jiguang Wan, Zhihu Tan, Changsheng Xie and Guokuan Li. WIPE: a Write-Optimized Learned Index for Persistent Memory (TACO 2023, CCF-A) Code

• Chen Ding, Jian Zhou, Jiguang Wan, Yiqin Xiong, Sicen Li, Shuning Chen, Hanyang Liu, Liu Tang, Ling Zhan, Kai Lu, Peng Xu. DComp: Efficient Offload of LSM-tree Compaction with Data Processing Units (ICPP 2023, CCF-B)

• Wei Tao, Shenglin He, Kai Lu, Xiaoyang Qu, Guokuan Li, Jiguang Wan, Jianzong Wang, Jing Xiao. Value-Driven Mixed-Precision Quantization for Patch-Based Inference on Microcontrollers (DATE 2023, CCF-B)

• Daping Li, Jiguang Wan, Jun Wang, Jian Zhou, Kai Lu, Peng Xu, Fei Wu and Changsheng Xie. Disperse Access Considered Energy Inefficiency in Intel Optane DC Persistent Memory Servers (ICDCS 2020, CCF-B) Code

• Ling Zhan, Kai Lu, Zhilong Cheng and Jiguang Wan. RangeKV: An Efficient Key-Value Store Based on Hybrid DRAM-NVM-SSD Storage Structure

• Ling Zhan, Kai Lu*, Yiqin Xiong, Jiguang Wan and Zixuan Yang. TrickleKV: A High-Performance Key-Value Store on Disaggregated Storage with Low Network Traffic

💬 Contact

• Email: kailu@hust.edu.cn
• Github: emperorlu; PDS-Lab