Changes:
- Removed GLM dependency and replaced it with native CUDA types: CUDA has good built-in types now to do what we need to do, and GLM's CUDA support was and is still experimental.
- Removed Thrust codepath: Performance was inferior, added complexity and maintenance not worth it.
- Moved Github Actions to latest CUDA version
- VoxFileWriter bump to latest version
- CUDA bump to 12.2
This zip file contains:
- 64-bit (Windows) RELEASE and DEBUG versions of cuda_voxelizer + required CUDA *.dlls
How to use:
Extract and run cuda_voxelizer.exe with valid program parameters