The code to our submitted paper to SIGMOD 2025 titled BLAEQ: Efficient on-GPU Spatial Range Query of Large-scale CFD Mesh using BLAS

The code is fully implemented in Python, including three small datasets for instant running.

The following requirements must be met:

Hardware:

An NVIDIA GPU supporting CUDA.

Environment:

-Python 3.x

-cupy

-numpy

-cupyx

-scipy

-pickle (for loading default data set)

-tqdm (for better display)

Simply drag all the .py files and the Data folder into a Python project and compile 'BLAEQ_Public.py'. The algorithm should start running with the default configuration. For details, plz refer to the comments within the code or our submitted paper.

Due to the restriction of file size. We were unable to release all the 9 datasets used in the paper onto github. Therefore, we only put the smallest 3 datasets into the \Data folder. The other files are released on Google Drive: https://drive.google.com/drive/folders/1rCV5jibO3h8tq00xPw3rj2sBz3GRKltZ?usp=drive_link

To use the other datasets, just put them into the \Data folder and change the 'case' variable in BLAEQ_Public.py with the appropriate file name.