grlee77
commented
2 years ago It looks like there are a number of CUDA implementations around, but I don't know if there are any with Python wrappers. I agree that it would be useful to provide.
various existing CUDA implementations
-
A BSD 3-clause licensed version in NVIDIA's CUDA samples.
-
A BSD 3-clause licensed version in NVlabs Intrinsics3D
-
Apache 2.0 licensed code in NVIDIAGameWorks kaolin voxelgrid.py
-
MIT-licensed 3rd party implementation in KinectFusionLib
-
BSD 2-clause licensed 3rd party implementation: marching-cubes-cuda
-
incompatibly licensed version in NVIDIA warp
-
incompatibly licensed version in NVlabs instant-ngp
-
incompatibly licensed "marching tetrahedrons" in NVlabs nvdiffrec (comments indicate that this is adapted from the kaolin code linked above)
Based on this user comment, seems like there may be some interest in a marching cube implementation on GPU.