Computer science · UC Riverside

Hongrui (Jack) Zhang

Ph.D. student · GPU computing & database systems

I study how specialized GPU hardware can make computation more efficient.

I am a Ph.D. student at the University of California, Riverside, advised by Hung-Wei Tseng in the ESCAL lab. My interests include sparse linear algebra, ray tracing accelerators, and database systems.

Research

Ongoing research

GPU-accelerated
database systems

I am exploring how CUDA, tensor, and ray tracing cores can work together to accelerate relational query processing. My goal is to keep data and intermediate results on the GPU while reducing data movement and the cost of constructing computational representations.

My current work examines algebraic representations of queries and opportunities to combine operations such as joins, grouping, and aggregation. I use TPC-H queries to study correctness and performance across the full computation pipeline.

RTSpMSpM · ISCA 2025

Ray tracing for
sparse matrix computation

In RTSpMSpM, I developed an approach that maps sparse matrix multiplication to ray tracing operations, using NVIDIA RT Cores for computation beyond graphics. I implemented the approach with OptiX and CUDA and integrated it into a graph neural network training framework.

Read the publication

Publications

Background

Education

Ph.D. in Computer Science
University of California, Riverside
– present
B.S. in Computer Engineering
University of California, San Diego
–

Earlier work

In 2023, I worked on machine learning for music, including genre-conditioned generation with PyTorch, genre classification using mel spectrograms, and audio compression and denoising with autoencoders.

Technical skills

Python, C++, C, Bash; CUDA, NVIDIA OptiX, cuSPARSE, PyTorch, NumPy; Linux, Git, and Docker.

Curriculum vitae

View CV PDF Download CV