Multi-block-LBM-GPU
Please refer the paper: Zhang, Ya, Guang Pan, and Qiaogao Huang. "ICCM2016: The implementation of two-dimensional multi-block lattice Boltzmann method on GPU." International Journal of Computational Methods 16, no. 05 (2019): 1840002. DOI: 10.1142/s0219876218400029
How to compile (assume cuda is installed properly, https://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/index.html https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html) nvcc -lcusparse -lcurand -lcublas Multi-block-LBM.cu -o test OR Visual Studio on Windows 1. Add the above libraries (cusparse.lib, curand.lib, cublas.lib) at Project Property Pages -> Linker -> Input -> Additional Dependencies; 2. Build the project; 3. Please use Nsight to debug the code (https://developer.nvidia.com/nsight-visual-studio-edition).