Hailong Zou
I am a Ph.D. student in Computer Science and Technology at the School of Computer Science, Peking University. At the Intelligence Fusion Laboratory (IF-Lab), I am advised by Prof. Xiang Chen. I work on efficient AI systems, with a focus on on-device foundation-model inference and hardware-software co-design for heterogeneous, resource-constrained platforms.
I received my M.Eng. in Circuits and Systems from the Institute of Microelectronics, Chinese Academy of Sciences, and my B.Eng. in Electronic Information Engineering from Wuhan University.
Research Interests
Efficient AI Systems
Efficient inference for foundation models, including LLM and MoE systems, caching, and model-system techniques that reduce memory and computation.
Hardware-Software Co-Design
Architecture-aware optimization across models, compilers, runtimes, and heterogeneous CPU, NPU, and accelerator platforms.
On-device & Edge AI
On-device deployment and edge-cloud collaboration under practical memory, storage, bandwidth, and latency constraints.
Selected Research
BigMoMo
BigMoMo uses speculative decoding to decouple expert movement from single-token execution, enabling weight reuse, contiguous flash reads, and transfer-compute overlap. Across four MoE models and five benchmarks on two mobile platforms, it achieves mean decoding speedups of 4.83× over on-demand offloading and 1.82× over the best speculative MoE baseline, while supporting models with up to 80B parameters.
Recent Publications
BigMoMo: Efficient Inference of Large-Scale MoE with Speculative Decoding on Mobile Devices
When and Where to Look: Adaptive Visual Evidence Scheduling for Efficient Long Video Understanding
EchoCache: Energy-Guided Cross-Modal Caching for Efficient Audio-Driven Video Generation
Proceedings of the 34th ACM International Conference on Multimedia (ACM MM), 2026. arXiv
MoECa: Aligning Feature Reuse with Expert Decomposition in Diffusion Transformers
Proceedings of the 34th ACM International Conference on Multimedia (ACM MM), 2026. arXiv
GOMA: Geometrically Optimal Mapping via Analytical Modeling for Spatial Accelerators
IEEE Transactions on Computers (TC), 2026. arXiv
Research & Industry Experience
Intelligence Fusion Laboratory (IF-Lab), Peking University
Ph.D. Student · Advisor: Prof. Xiang Chen
Research on efficient AI systems, on-device foundation-model inference, hardware-software co-design, and heterogeneous acceleration.
Institute of Microelectronics, Chinese Academy of Sciences
Core Team Member · Energy-Efficient Neuromorphic Processing Chip Project
Developed a PCIe host-to-DDR data-transfer module and optimized YOLOv3-SpikeformerV2 through pretraining and quantization-aware training.
Empyrean Technology, Chengdu
Software Engineering Intern · BTG Team
Contributed to pyAether by exposing core C++ capabilities of the Aether EDA platform through Python APIs and validating interface correctness and coverage.
Education
Peking University
Ph.D. in Computer Science and Technology, School of Computer Science.
Intelligence Fusion Laboratory (IF-Lab), Advisor: Prof. Xiang Chen.
University of Chinese Academy of Sciences
M.Eng. in Circuits and Systems, Institute of Microelectronics, Chinese Academy of Sciences.
Wuhan University
B.Eng. in Electronic Information Engineering, Excellence Engineer Program.
Honors & Awards
- UCAS Outstanding Student, 2023-2024
- UCAS Outstanding Youth League Member, 2023-2024
- Hubei Undergraduate Electronic Design Contest (TI Cup), Special Prize, 2020-2021
- Wuhan University Outstanding Student & Third-Class Scholarship, 2020-2021
- Wuhan University Social Engagement Award, 2021-2022
- Outstanding Activity Organizer, China Foundation for Poverty Alleviation, 2021-2022
Academic Service & Activities
Conference Reviewer
Conference on Neural Information Processing Systems (NeurIPS), IEEE/ACM International Conference on Computer-Aided Design (ICCAD), and IEEE International Conference on Computer Design (ICCD).
Leadership & Service
Former Director of External Relations, Wuhan University Youth Volunteers Association.
Photography
Contract Photographer at Visual China Group (VCG) and Contract Contributor at Tuchong.
Photography
Selected photographic works. Click any image to view it in full.
