Will Open3D automatically use cuda to speed up?

[etiwal]

Is Open3D automatically choosing the most performant method for building KDTrees etc.?

I'm using the icp registration module and i can't see any speedup from v.12.0 to v.13.0

[yxlao]

The new CUDA-accelerated stuff is in the open3d::core and open3d::t namespace, where t stands for Tensor.
Check out https://github.com/intel-isl/Open3D/blob/master/cpp/open3d/t/pipelines/registration/Registration.h

[etiwal]

Thanks for your answer! I'm using the Python API.

Should i first move the target point cloud to the GPU and then apply ICP registration on this? If so how would i move it? Similar to pytorchs point_cloud.to_device()?

[etiwal]

Dear @yxlao

Do you have an update on how to use the ICP registration on GPU through the Python API?

[ssheorey]

http://www.open3d.org/docs/latest/tutorial/t_pipelines/t_icp_registration.html#ICP-registration