Scholar
Liangyu Zhao
Google Scholar ID: xFz1700AAAAJ
CS PhD Student, University of Washington - Seattle
Citations & Impact
Citations
475
H-index
7
i10-index
7
Publications
11
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- Publications:
- - ForestColl: Throughput-Optimal Collective Communications on Heterogeneous Network Fabrics
- - FLASH: Fast All-to-All Communication in GPU Clusters
- - NanoFlow: Towards Optimal Large Language Model Serving Throughput
- - Efficient Direct-Connect Topologies for Collective Communications
- - Rethinking Machine Learning Collective Communication as a Multi-Commodity Flow Problem
- - Efficient all-to-all Collective Communication Schedules for Direct-connect Topologies
- - AutoLRS: Automatic Learning-Rate Schedule by Bayesian Optimization on the Fly
- - Nexus: A GPU Cluster Engine for Accelerating DNN-Based Video Analysis
Research Experience
- Will join Meta Superintelligence Lab (MSL) Infra team as a research scientist intern in June 2025; will join NVIDIA Applied Deep Learning Research (ADLR) group as a research intern in March 2024.
Education
- Currently a fourth-year Ph.D. student in the Paul G. Allen School of Computer Science & Engineering at the University of Washington, advised by Prof. Arvind Krishnamurthy. Completed undergraduate studies at the University of Washington, earning a B.S. in Computer Science and a B.S. in Applied & Computational Mathematical Sciences (Discrete Math and Algorithms).
Background
- Research interests include machine learning systems, distributed systems, and collective communications. Broadly speaking, interested in formulating and solving mathematical problems in computer systems and networking.
Co-authors