Ph.D. Student · Computer Science
Bowen Cui
I research AI infrastructure, GPU kernel performance, and LLM-driven optimization of high-performance computing workloads.
I am a Computer Science Ph.D. student at George Mason University, advised by Professor Keren Zhou. My research combines performance benchmarking, compiler and hardware analysis, and reliable code optimization across GPUs and HPC applications. I previously earned an M.A. in Statistics from Washington University in St. Louis and a B.S. in Applied Mathematics from Xi'an Jiaotong-Liverpool University.
Research Interests
GPU Kernels & Accelerators
Tile-based programming models, compiler lowering, profiling, and cross-architecture performance.
LLMs for HPC Optimization
Benchmarking, performance-guided optimization agents, correctness, and reproducible evaluation.
AI Infrastructure
Scalable benchmark infrastructure for real scientific applications, from GPU kernels to multi-GPU workloads.
GitHub Activity
News
Oct 2026
Our work on physics-informed graph convolutional networks for snow water equivalent estimation was accepted by Earth and Space Science.
Sep 2026
HPC-Bench was accepted to the NeurIPS 2026 Evaluations & Datasets Track (poster).
Sep 2026
TileBench, our benchmark of Triton and cuTile programming models, was accepted to AACL 2026.
Sep 2025
Presented our study of LLM-driven HPC code optimization at ICPP Companion 2025.
Aug 2024
Started my Computer Science Ph.D. at George Mason University, advised by Prof. Keren Zhou.
Selected Publications
Estimating Regional Snow Water Equivalent Using a Physics-Informed Graph Convolution Network Model
Wei-Ting Hung, Ziheng Sun, Bowen Cui, Keren Zhou, Nicoleta Cristea, Justin Pflug, Annie Burgess, Daniel Tong
Earth and Space Science2026
TileBench: A Controlled Benchmark for Performance Evaluation and Bottleneck Diagnosis of Tile-Based Programming Models
Bowen Cui*, Zhongchun Zhou*, Hao Wu, Tejas Ramesh, Junyu Yin, Jialiang Gu, Keren Zhou (* equal contribution)
AACL 20262026
HPC-Bench: A Comprehensive Benchmark for High Performance Computing Codes
View all publications →
Explore research projects →
Bowen Cui, Junyu Yin, Tejas Ramesh, Leo Lim, Oscar Hernandez, Keren Zhou
NeurIPS 2026 E&D2026