XIAOBO ZHOU        

Prof. Zhou’s research is centered on advancing system support for data-intensive applications, particularly in the realms of cloud computing and systems for machine learning. He focuses on optimizing performance, resource efficiency, and system reliability, all of which are critical for scaling modern computational workloads.

He served as the chair of the IEEE Technical Committee on Distributed Processing (2020-2023).

Research Interests

• Cloud Computing and Virtulization
• Systems for Machine Learning
• High-Performance Distributed Computing
• Memory Systems and OS

PhD Student Opportunities

* Prof. Zhou has supervised over a dozen Ph.D. students, with graduates securing tenure-track faculty positions at research-oriented academic institutions. He is looking for self-motivated PhD students who have solid background and interests in Operating System, Computer Architecture, Parallel and Distributed Computing, and Systems ML-related fields. Please send him an email with your CV.  

Recent Publications

• [OSDI26] Unleash All Cores: Scalable Asymmetric-aware Scheduling for DNN Inference on Mobile CPUs, USENIX OSDI, 2026
• [PPoPP26] JanusQuant: Accurate and Efficient 2-bit KV Cache Quantization for Long-Context Inference, ACM PPoPP, 2026
• [PPoPP25] Harnessing Inter-GPU Shared Memory for Seamless MoE Communication-Computation Fusion, ACM PPoPP, 2025
• [SC25] MXBLAS: Accelerating 8-bit Deep Learning with a Unified Micro-Scaled GEMM Library, ACM/IEEE SC, 2025
• [SC25] HyTiS: Hybrid Tile Scheduling for GPU GEMM with Enhanced Wave Utilization and Cache Locality, ACM/IEEE SC, 2025
• [ATC25] Sparse Matrix-Matrix Multiplication on Tensor Cores with Asynchronous and Balanced Kernel Optimization, USENIX ATC, 2025
• [SC24] Scaling New Heights: Transformative Cross-GPU Sampling for Training Billion-Edge Graphs, ACM/IEEE SC, 2024
• [SC24] MCFuser: High-Performance and Rapid Fusion of Memory-bound Compute-intensive Operators, ACM/IEEE SC, 2024
• [SC24] Accelerating Distributed DLRM Training with Optimized TT Decomposition and Micro-Batching, ACM/IEEE SC, 2024
• [ATC24] Expeditious High-Concurrency MicroVM SnapStart in Persistent Memory with an Augmented Hypervisor, USENIX ATC, 2024