Paper 'TCCL: Automated Collective Communication Tuning for Accelerating Distributed and Parallel DNN Training' accepted to NSDI 2025
Paper 'BigMac: A Light All-to-All Mixture-of-Experts Model Structure for Fast Training and Inference' accepted to AAAI 2025
Paper 'PolyBase: Adapting to Data Affinity Changes in Geo-Replicated Database via Row-Level Paxos-Group Affiliation Re-Assignment' accepted to VLDB 2025
Paper 'VPTQ: Extreme Low-bit Vector Post-Training Quantization for Large Language Models' accepted to EMNLP 2024 Main Conference
Paper 'Eliminating Data Processing Bottlenecks in GNN Training over Large Graphs via Two-level Feature Compression' accepted to VLDB 2024
Paper 'Cuber: Constraint-Guided Parallelization Plan Generation for Deep Learning Training' accepted to OSDI 2024
Paper 'Noctua: Towards Automated and Practical Fine-grained Consistency Analysis' published at EuroSys 2024
Supervised Chaoyi Ruan’s successful PhD defense titled 'Design and Optimization of Resource Scaling and High Availability for Cloud-Native Databases'