https://hgpu.org/?p=5753
Parallel Computing based on GPGPU using Compute Unified Device Architecture